From owner-structural-nmr@net.bio.net Thu Apr 07 23:00:00 1994 Path: biosci!biosci!not-for-mail From: root@net.bio.net (Operator) Newsgroups: bionet.structural-nmr Subject: test of bionet.structural-nmr Date: 8 Apr 1994 13:06:27 -0700 Organization: BIOSCI International Newsgroups for Biology Lines: 15 Distribution: bionet Message-ID: <2o4dg3$g23@net.bio.net> NNTP-Posting-Host: net.bio.net test of bionet.structural-nmr Please do not post to this newsgroup until the official announcement is out that both the group and the mail-to-news gateways are operational. Thank you. Sincerely, Dave Kristofferson BIOSCI/bionet Manager biosci-help@net.bio.net From owner-structural-nmr@net.bio.net Thu Apr 07 23:00:00 1994 Path: biosci!NET.BIO.NET!biosci-help From: biosci-help@NET.BIO.NET (BIOSCI Administrator) Newsgroups: bionet.structural-nmr Subject: test of str-nmr@net.bio.net Date: 8 Apr 1994 13:11:30 -0700 Organization: BIOSCI International Newsgroups for Molecular Biology Lines: 2 Sender: daemon@net.bio.net Distribution: bionet Message-ID: Reply-To: biosci-help@net.bio.net NNTP-Posting-Host: net.bio.net test of str-nmr@net.bio.net From owner-structural-nmr@net.bio.net Thu Apr 07 23:00:00 1994 Path: biosci!daresbury!not-for-mail From: BIOSCI Administrator Newsgroups: bionet.structural-nmr Subject: test of str-nmr@daresbury.ac.uk Date: 8 Apr 1994 21:14:11 +0100 Lines: 3 Sender: daemon@mserv1.dl.ac.uk Distribution: bionet Message-ID: <2o4duj$o4u@mserv1.dl.ac.uk> Reply-To: biosci-help@net.bio.net Original-To: str-nmr@dl.ac.uk test of str-nmr@daresbury.ac.uk From owner-structural-nmr@net.bio.net Thu Apr 07 23:00:00 1994 Path: biosci!NET.BIO.NET!biosci-help From: biosci-help@NET.BIO.NET (BIOSCI Administrator) Newsgroups: bionet.structural-nmr Subject: Welcome to the STRUCTURAL-NMR/bionet.structural-nmr newsgroup Date: 8 Apr 1994 16:49:51 -0700 Organization: BIOSCI International Newsgroups for Molecular Biology Lines: 183 Sender: daemon@net.bio.net Distribution: bionet Message-ID: Reply-To: biosci-help@net.bio.net NNTP-Posting-Host: net.bio.net PLEASE READ THE FOLLOWING CAREFULLY AND SAVE FOR FUTURE REFERENCE! The STRUCTURAL-NMR/bionet.structural-nmr newsgroup is ready for operation. The former address nmr-str@net.bio.net for the prototype newsgroup NMR-STR has been forwarded to the new str-nmr@net.bio.net address so that both addresses will route mail to the same mailing list and USENET newsgroup. IMPORTANT!! - For the next week all of the former members of the nmr-str mailing list remain on the str-nmr mailing list at net.bio.net. Next week I will transfer all **European, African, and Central Asian** e-mail addresses to our UK distribution site at Daresbury. You can contact them at biosci@daresbury.ac.uk. FIRST, HOWEVER, ANYONE WHO HAS ACCESS TO THE USENET NEWSGROUP bionet.structural-nmr SHOULD CANCEL THEIR E-MAIL SUBSCRIPTION and use news instead. News is much more convenient than e-mail. To do this, address a message to biosci-server@net.bio.net Leave the Subject: line blank and then include in the body of your message the line unsubscribe str-nmr NOTE that you MUST use str-nmr in this message; nmr-str WILL NOT WORK!! PLEASE NOTE that many USENET sites do not allow automatic creation of new USENET groups!!! If you do not see bionet.structural-nmr in your newsreader within another day or two, tell your systems administrator to get on the ball and act on our "newgroup" message to enable the group at your site. We have already done several tests and are certain that the group is currently propagating around the network. Now for the usual info about newsgroup mechanics: Subscribing to this list: ------------------------- First if you received this message you are ALREADY SUBSCRIBED to the mailing list, so you need NOT do anything further (unless you want to cancel because you intend to use news instead). If you have colleagues who would like to sign up for e-mail distribution, they should proceed as follows. IF YOU ARE LOCATED IN EUROPE, AFRICA, OR CENTRAL ASIA: please contact biosci@daresbury.ac.uk and request a subscription to STRUCTURAL-NMR. IF YOU ARE LOCATED IN THE AMERICAS OR THE PACIFIC RIM: log in to the computer account in which you would like to receive mail (not an account that you use infrequently) and send a mail message to the Internet address biosci-server@net.bio.net Leave the Subject: line of the message blank and enter the following line into the body of the mail message: subscribe str-nmr This message will be automatically read by our computer and your e-mail address will be extracted from the mail header and added to the list. Canceling your subscription: ---------------------------- IF YOU ARE LOCATED IN EUROPE, AFRICA, OR CENTRAL ASIA: please contact biosci@daresbury.ac.uk and request that your subscription to STRCUTURAL-NMR be canceled. IF YOU ARE LOCATED IN THE AMERICAS OR THE PACIFIC RIM: send a message to biosci-server@net.bio.net exactly as described above for subscribing except include the text unsubscribe str-nmr in the body of the message. Please be sure to send the message from the account whose address matches the one on the list. IF YOU HAVE A PROBLEM: ---------------------- Please send a message to one of the following addresses depending upon your location Address Location ------- -------- biosci@daresbury.ac.uk Europe, Africa, and Central Asia biosci-help@net.bio.net Americas and the Pacific Rim and someone on the staff will help you. PLEASE DO NOT send mail to our personal e-mail addresses as this will delay a response to your request for help. How to post a message to the group: ----------------------------------- If you use news, simply post a message into bionet.structural-nmr. Be sure to set your "distribution" to either bionet or world or else the message might not leave your site!! To post by e-mail, mail your message to one of the following addresses depending upon your location: Posting Address Location --------------- -------- str-nmr@daresbury.ac.uk Europe, Africa, and Central Asia str-nmr@net.bio.net Americas and the Pacific Rim and your message will be distributed automatically to everyone on the list and the USENET newsgroup. There is no editorial intervention. PLEASE DO NOT SEND SUBSCRIPTION REQUESTS TO THE POSTING ADDRESSES as you will bother everyone on the newsgroup!!! How to reply to a message on the group: --------------------------------------- If you are using a newsreader, simply use the reply or follow-up command on your newsreader (these vary from program to program) to send either private or public replies. If you are using e-mail, replies to messages that you receive will *NOT* be automatically returned to the group. This is the standard for Internet mailing lists as opposed to BITNET LISTSERVs which often send all replies back to everyone. You must be certain that your reply contains either of the two newsgroup posting addresses above in your message header if you want to share it with everyone on the group. Otherwise in most cases your reply may go back to only the original poster of the message to which you are replying. ALWAYS be certain that you examine the address on your messages before you send them!!! Once a message is sent there is no way to cancel it or bring it back!!! Some non-Internet compliant mail systems may attempt to send replies to our error-trapping address called BIOSCI-REQUEST. If yours does this, please be sure to readdress your message to str-nmr@net.bio.net or str-nmr@daresbury.ac.uk if you want to send it to the newsgroup. How to look at archives of the list: ------------------------------------ Archives for STRUCTURAL-NMR/bionet.structural-nmr are kept in the anonymous FTP account at net.bio.net [134.172.2.69]. Look in the directory pub/BIOSCI/STRUCTURAL-NMR for posting archives. Each file is assigned a date such as 9312 for December 1993. Please note that ours is a UNIX system and all file and directory names are case-sensitive, i.e., upper case file names are different from lower case names. You can also access these same files via Gopher if you start a gopher session using net.bio.net as your gopher server. Gopher also allows you to view the individual messages within each monthly archive file. The files are in the STRUCTURAL-NMR directory. Postings to bionet.structural-nmr are also WAIS indexed and can be searched via either gopher or WAIS at our site. In gopher the option at net.bio.net is "Search Bionet USENET Articles" and in WAIS one should use the WAIS source biosci.src. This is a WAIS index of all BIOSCI/bionet messages including this newsgroup. Please see the BIOSCI FAQ for details. The FAQ can be requested from biosci-help@net.bio.net. Once again, if you have any administrative questions that require personal assistance, please address them to biosci-help@net.bio.net in the U.S. or biosci@daresbury.ac.uk in the UK. Best wishes for a successful newsgroup! Sincerely, Dave Kristofferson BIOSCI/bionet Manager biosci-help@net.bio.net From owner-structural-nmr@net.bio.net Thu Apr 07 23:00:00 1994 Path: biosci!biosci!not-for-mail From: kristoff@net.bio.net (David Kristofferson) Newsgroups: bionet.structural-nmr Subject: test 2 of bionet.structural-nmr Date: 8 Apr 1994 13:29:51 -0700 Organization: BIOSCI International Newsgroups for Biology Lines: 1 Distribution: bionet Message-ID: <2o4erv$hhr@net.bio.net> NNTP-Posting-Host: net.bio.net test 2 of bionet.structural-nmr From owner-structural-nmr@net.bio.net Sun Apr 10 23:00:00 1994 Newsgroups: bionet.structural-nmr Path: biosci!bloom-beacon.mit.edu!usc!cs.utexas.edu!howland.reston.ans.net!torn!utnut!oci!greig From: greig@tris.oci.utoronto.ca (David Iain Greig) Subject: Welcome, welcome... Message-ID: Followup-To: bionet.structural-nmr Sender: greig@tris (David Iain Greig) Organization: Structural Biology, Ontario Cancer Institute Date: Mon, 11 Apr 1994 15:15:09 GMT Lines: 32 Since we've had this newsgroup up and running at our site for several days now, and yet had no posts, I thought I would make the obligatory attempt at making the first post. I am currently finishing up my M.Sc. in Biophysics at the University of Toronto working on the NMR-derived structure of a tetradecameric DNA duplex d(GGGCTATAAAAGGG)-d(CCCTTTTATAGCCC). In writing up my thesis, I am forced to conclude that 1H NMR of ordinary B-form DNA is of limited utility in discerning fine structure present in the duplex. I do see certain patterns in the NOE spectrum, such as very weak (1',8) sequential peaks in the second T-A step in *both* strands, indicating either a kink or propellor twist or the like, but the RMS deviation in my rMD/SA structures (using X-PLOR 3.1) is over an angstrom, even if I do get minimal violations of my NOE constraints. And the helix is underwound. Sigh. I am pretty much in agreement with David Wemmer(*) on this question of the utility of NMR on duplex oligos. Anyone else care to comment? (*) Wemmer, D. (1991) "The applicability of NMR methods to solution structure of nucleic acids" _Curr. Op. Struct. Biol_, 1:452-458. Sincerely Dave -------------------------------------------------------------------------- David Iain Greig greig@oci.utoronto.ca Dept. of Medical Biophysics RIPEM/PGP Public Keys available on request University of Toronto "A clean bathroom makes a quiet mind" Ontario Cancer Institute/Princess Margaret Hospital Kibo Number 1 -------------------------------------------------------------------------- From owner-structural-nmr@net.bio.net Mon Apr 11 23:00:00 1994 Newsgroups: bionet.structural-nmr Path: biosci!bloom-beacon.mit.edu!usc!cs.utexas.edu!howland.reston.ans.net!EU.net!dkuug!uts!unidhp.uni-c.dk!carlmk From: carlmk@unidhp.uni-c.dk (Mogens Kjaer) Subject: What is this? Organization: UNI-C, Danish Computing Centre for Research and Education Date: Tue, 12 Apr 1994 09:30:21 GMT Message-ID: <1994Apr12.093021.8263@uts.uni-c.dk> X-Newsreader: TIN [version 1.2 PL1] Sender: news@uts.uni-c.dk (News) Nntp-Posting-Host: unidhp.uni-c.dk Lines: 17 This newsgroup popped up im my newsreader a few days ago, but noone have posted anything so far - is there something wrong, or are there nobody out there working with structural NMR? Mogens -- Mogens Kjaer Pronto Software - Development and Distribution Carlsberg Research Center Gamle Carlsberg Vej 10 DK-2500 Valby Denmark Phone: + 45 33 27 53 25 Fax: + 45 33 27 47 08 Email: carlmk@unidhp.uni-c.dk From owner-structural-nmr@net.bio.net Mon Apr 11 23:00:00 1994 Newsgroups: bionet.structural-nmr Path: biosci!bloom-beacon.mit.edu!usc!howland.reston.ans.net!EU.net!dkuug!uts!news From: cc@charon.dfh.dk Subject: Re: What is this? Reply-To: cc@charon.dfh.dk Organization: Royal Danish School of Pharmacy Date: Tue, 12 Apr 1994 12:10:38 GMT Message-ID: <1994Apr12.121038.14608@uts.uni-c.dk> Lines: 29 X-Newsreader: IBM NewsReader/2 v1.00 References: <1994Apr12.093021.8263@uts.uni-c.dk> Sender: news@uts.uni-c.dk (News) Nntp-Posting-Host: cornetto.dfh.dk Lines: 30 In <1994Apr12.093021.8263@uts.uni-c.dk>, carlmk@unidhp.uni-c.dk (Mogens Kjaer) writes: >This newsgroup popped up im my newsreader a few days ago, but noone have >posted anything so far - is there something wrong, or are there nobody >out there working with structural NMR? > >Mogens > >-- >Mogens Kjaer >Pronto Software - Development and Distribution >Carlsberg Research Center >Gamle Carlsberg Vej 10 >DK-2500 Valby >Denmark >Phone: + 45 33 27 53 25 >Fax: + 45 33 27 47 08 >Email: carlmk@unidhp.uni-c.dk > The thing is, this group is brand new and probably not widely enough advertised, but the answer is that at another person works with structural NMR 'round here... Claus Cornett Institute for Analytical and Pharmaceutical Chemistry The Royal Danish School of Pharmacy Universitetsparken 2 DK 2100 Copenhagen Fax. +45 35375744 From owner-structural-nmr@net.bio.net Tue Apr 12 23:00:00 1994 Path: biosci!bloom-beacon.mit.edu!senator-bedfellow.mit.edu!jrwill From: jrwill@oxytricha.mit.edu (Jamie Williamson) Newsgroups: bionet.structural-nmr Subject: Re: duplex DNA NMR Date: 13 Apr 1994 02:01:30 GMT Organization: Massachvsetts Institvte of Technology Lines: 53 Distribution: world Message-ID: <1994Apr12.214434@oxytricha.mit.edu> References: NNTP-Posting-Host: oxytricha.mit.edu Originator: jrwill@oxytricha.mit.edu David Iain Greig writes: >I am currently finishing up my M.Sc. in Biophysics at the University of Toronto >working on the NMR-derived structure of a tetradecameric DNA duplex >d(GGGCTATAAAAGGG)-d(CCCTTTTATAGCCC). In writing up my thesis, I am forced >to conclude that 1H NMR of ordinary B-form DNA is of limited utility in >discerning fine structure present in the duplex. >I do see certain patterns in the NOE spectrum, such as very weak (1',8) >sequential peaks in the second T-A step in *both* strands, indicating either >a kink or propellor twist or the like, but the RMS deviation in my rMD/SA >structures (using X-PLOR 3.1) is over an angstrom, even if I do get >minimal violations of my NOE constraints. And the helix is underwound. >Sigh. >I am pretty much in agreement with David Wemmer(*) on this question of the >utility of NMR on duplex oligos. Anyone else care to comment? >(*) Wemmer, D. (1991) "The applicability of NMR methods to solution structure >of nucleic acids" _Curr. Op. Struct. Biol_, 1:452-458. I care to comment! I sympathize with you completely, because when I wrote my thesis in 1988, I came to the same conclusion! It is apparently very difficult to really nail down local structural variations using the NOE data we can usually get. Pardi wrote a paper (JMB in 1991 or 1992 I think) that used synthetic exact data to determine a structure. I believe that he found that glycosidic conformation and pucker were well determined. However, the values for other torsion angles usually varied completely over the range of crystallographically observed values. You didn't mention if you had used relaxation matrix methods or not. They are reported to help some. I personally have no experience with those methods because I bagged it before they came around. The problem is that nucleic acids, in particular DNA, have a very different topology for the spins compared to proteins. There are very few interresidue NOEs that connect nucleotides that are not adjacent. We get a very local picture of what is going on. Anyone care to vouch for the efficacy of relaxation matrix stuff? Jamie Williamson Dept. Chemistry MIT From owner-structural-nmr@net.bio.net Tue Apr 12 23:00:00 1994 Path: biosci!IRIS.BCHEM.KSU.EDU!drm From: drm@IRIS.BCHEM.KSU.EDU (Dave Manning) Newsgroups: bionet.structural-nmr Subject: Varian NMR & UPS Date: 12 Apr 1994 18:10:39 -0700 Organization: BIOSCI International Newsgroups for Molecular Biology Lines: 22 Sender: daemon@net.bio.net Distribution: bionet Message-ID: <9404130057.AA02614@iris.bchem.ksu.edu> NNTP-Posting-Host: net.bio.net The "purchasing procedure" has been undertaken for obtaining an uninterruptible power system UPS (~4.3 KVa) for a use in our NMR center. This unit will be used for the following equipment: Varian Unity Plus (500 MHz) SGI Indigo^2 XZ The purchase request was made, using specifications from Best Power Technology (their Ferrups Model FD4.3KVa/3KW). A vendor has offered a solution which includes a 4.5 KVa system manufactured by Computer Power Inc (High Bridge, NJ). The Computersave Mark II is of the ferroresonant technology. Is there any Varian NMR equipment out there being used with these Computer Power Inc systems? (Sorry, this isn't *real* structural!!) -- Dave Manning / Kansas State Univ Bitnet: dave2@ksuvm Phone: (913) 532-6125 Biochemistry Internet: drm@iris.bchem.ksu.edu From owner-structural-nmr@net.bio.net Tue Apr 12 23:00:00 1994 Path: biosci!bloom-beacon.mit.edu!senator-bedfellow.mit.edu!jrwill From: jrwill@oxytricha.mit.edu (Jamie Williamson) Newsgroups: bionet.structural-nmr Subject: DQ relaxation Date: 13 Apr 1994 01:47:11 GMT Organization: Massachvsetts Institvte of Technology Lines: 19 Distribution: world Message-ID: <1994Apr12.213453@oxytricha.mit.edu> NNTP-Posting-Host: oxytricha.mit.edu Originator: jrwill@oxytricha.mit.edu Hello: Does anyone know the formula for T2 for a homonuclear double quantum coherence? I suppose there should be spectral densities J(0), J(w), and J(2w) just like for regular T2, but the coefficients are probably different. Which term is biggest, especially for large molecules ? If J(0) dominates, the DQ T2 should be no worse than the SQ T2. Yes? No? Jamie Williamson Dept. Chemistry MIT From owner-structural-nmr@net.bio.net Tue Apr 12 23:00:00 1994 Path: biosci!psipsy.uct.ac.za!JACKSON From: JACKSON@psipsy.uct.ac.za ("Graham E Jackson") Newsgroups: bionet.structural-nmr Subject: Geminal proton distances Date: 12 Apr 1994 23:31:42 -0700 Organization: BIOSCI International Newsgroups for Molecular Biology Lines: 27 Sender: daemon@net.bio.net Distribution: bionet Message-ID: NNTP-Posting-Host: net.bio.net Hi, I am trying to use nOe data to constrain the interproton distances in a decapeptide but am having difficulty in scaling my data. If you use the nOe between two beta protons what do you use as the interproton distance. The crystallographers tell me it is 1.76A yet the sum of the van der Waals radii is 2.4A. The first value gives much to short interproton distances for all the other protons but I am loath to use 2.4A just because it works. A second problem is how do you integrate your nOe cross peaks. Do you 'fold' the data and what do you do with overlapping peaks? At the moment I am using the peak integration routine given in the Varian VnmrS software. Thanks for am assistance Graham E Jackson Graham E Jackson Department of Chemistry University of Cape Town Cape Town South Africa Fax: (027 21) 6503788 Tel: (027 21) 6502531 E-mail: jackson@psipsy.uct.ac.za From owner-structural-nmr@net.bio.net Tue Apr 12 23:00:00 1994 Path: biosci!sneezy.abbott.com!rob From: rob@sneezy.abbott.com (Robert Powers Meadows) Newsgroups: bionet.structural-nmr Subject: DNA structures by NMR Date: 13 Apr 1994 05:12:13 -0700 Organization: BIOSCI International Newsgroups for Molecular Biology Lines: 44 Sender: daemon@net.bio.net Distribution: bionet Message-ID: <9404131212.AA20059@sneezy.abbott.com> NNTP-Posting-Host: net.bio.net David Iain Greig greig@oci.utoronto.ca writes: > > I am currently finishing up my M.Sc. in Biophysics at the University of Toronto > working on the NMR-derived structure of a tetradecameric DNA duplex > d(GGGCTATAAAAGGG)-d(CCCTTTTATAGCCC). In writing up my thesis, I am forced > to conclude that 1H NMR of ordinary B-form DNA is of limited utility in > discerning fine structure present in the duplex. > During my Ph.D. thesis work, I too was mired in the MORASS (:-) of NMR structure determination of oligonucleotides. In order to get any fine structure, the general consenus was that you needed to account somehow for spin diffusion in order to get more "accurate distances." Since one is more confident in the distances, error bars may be reduced on the restraints. Thus, the precision and accuracy of the final structures may be increased (or so the story goes). This is especially true for essentially canonical oligonucleotides where most of the structural information comes from long distance NOES. Thus, I coded up a complete relaxation matrix analysis program. VERY similar to Rob Kapteins IRMA. It's available via anonymous ftp from dggpi2@chem.purdue.edu in the /pub/morass directory. Give it a whirl, it may help. But get some aspirin first. > I am pretty much in agreement with David Wemmer(*) on this question of the > utility of NMR on duplex oligos. Anyone else care to comment? Ahhh, well to be honest, I'm now at Abbott Labs doing NMR determination of proteins and protein complexes. No DNA by itself in solution. Let me say this, IMHO to a great degree both Wemmer (re the DNA question) and Clore (re the quantification question) were right. Although on both questions there is still lots of vociferous debate. Good luck Rob Meadows, Ph.D. Department of Computational and Molecular Sciences Abbott Labs Abbott Park, IL 60064 The views expressed are my own and do not reflect those of my employer. From owner-structural-nmr@net.bio.net Tue Apr 12 23:00:00 1994 Path: biosci!BIOC01.UTHSCSA.EDU!raman From: raman@BIOC01.UTHSCSA.EDU (C.S.RAMAN) Newsgroups: bionet.structural-nmr Subject: NMR - educational materials.. Date: 12 Apr 1994 21:11:58 -0700 Organization: BIOSCI International Newsgroups for Molecular Biology Lines: 33 Sender: daemon@net.bio.net Distribution: bionet Message-ID: <9404130413.AA11200@bioc01.uthscsa.edu> NNTP-Posting-Host: net.bio.net Dear Str-NMR readers: Thanks to everyone who voted YES for creating this newsgroup. I am glad to see postings on this newsgroup and encourage others to participate without inhibition. One of the ideas I entertained in my earlier posting was the availability of "Educational Materials" related to Structural NMR. I have received a large number of e-mail supporting this initiative from both experts and newcomers to this field. So, if you would like to contribute please send me a note on the nature of the material. Suitable items will ultimately end up on our "Structural-NMR Gopher" for easy downloading. Both electronic and hardcopy versions are welcome! Also, I am in the process of creating a Structural-NMR user directory which will contain names, addresses, and email of all scientists in this community. I will be distributing a template in the next couple of weeks that will provide additional details. Cheers -raman -- C.S.Raman UNIX Programming & Administration SPARC & SGI Systems raman@bioc01.uthscsa.edu - INTERNET Department of Biochemistry raman@mintaka.chpc.utexas.edu - CHPC UTHSCSA c.raman@launchpad.unc.edu 7703 Floyd Curl Dr. (210) 567-6623 [Tel] San Antonio, TX 78284-7760 (210) 567-6595 [Fax] ****************************************************************************** If a man's wit be wandering, let him study the Mathematics -Francis Bacon ****************************************************************************** From owner-structural-nmr@net.bio.net Wed Apr 13 23:00:00 1994 Path: biosci!TOME.CBS.UNIV-MONTP1.FR!jrene From: jrene@TOME.CBS.UNIV-MONTP1.FR (Jean Rene Alattia) Newsgroups: bionet.structural-nmr Subject: Moldyn Date: 14 Apr 1994 08:13:34 -0700 Organization: BIOSCI International Newsgroups for Molecular Biology Lines: 15 Sender: daemon@net.bio.net Distribution: bionet Message-ID: <199404141513.IAA02143@net.bio.net> NNTP-Posting-Host: net.bio.net Hello Can someboy tell me where I can get the Moldyn software for NMR relaxation data optimization. It is an old program from the early 80s , but I saw in a recent article that new things have been added to it. ######################################################################### # Jean-Rene ALATTIA # # # Centre de Biochimie Structurale # TEL: (33) 67 04 34 33 # # Faculte de Phrmacie # FAX: (33) 67 52 96 23 # # 15, Avenue Charles Flahault # # # 34060 Montpellier Cedex 1 # jrene@cbs.univ-montp1.fr # ######################################################################### From owner-structural-nmr@net.bio.net Wed Apr 13 23:00:00 1994 Path: biosci!mcs.anl.gov!maltsev From: maltsev@mcs.anl.gov (Natalie Maltsev) Newsgroups: bionet.structural-nmr Subject: a workshop announcement Date: 14 Apr 1994 10:47:37 -0700 Organization: BIOSCI International Newsgroups for Molecular Biology Lines: 163 Sender: daemon@net.bio.net Distribution: bionet Message-ID: <199404141744.MAA20529@juju> NNTP-Posting-Host: net.bio.net ------------------------------------------------------------------- Dear reader: Please bring this notice to the attention of persons you think might wish to participate. Workshop on Integration of Biochemical Databases June 8 - June 10, 1994 Pushchino, Russia (Registration deadline May 12) We invite all interested researchers to visit this major center of biomedical research in Russia and the former Soviet Union. The workshop will give you the opportunity to informally discuss how to integrate databases relevant to molecular biology and medicine. Several leading biological database experts from outside Russia will be present, and many interesting (and probably unfamiliar) Russian database projects will be demonstrated by their representatives. The workshop will be hosted by Professor Evgeni E. Selkov and coworkers, and is arranged in part by Dr. Ross Overbeek et al. at Argonne National Laboratory, Illinois, U.S.A. We have scheduled the following events: o Overview of EMP (Enzymes and Metabolic Pathways database). EMP contains the encoding of the most important articles (approx. 10,000) describing enzymes, phylogenetically arranged. It also contains in graphical form most of the known metabolic pathways for all organisms. We consider EMP the world's leading database of its kind. All participants will receive a free academic copy of this database (see details below). o For the first time, a presentation of a broad selection of Russian / FSU databases. There will be poster sessions and demonstrations on PC computers. During this winter we have collected brief descriptions and invited representatives from the groups who created (in our judgment) the most interesting of these. In Russia there exist substantial data collections ranging from minor ones done by single experts to very large collections of for example medicinal properties of compounds, 3D structural data, sequence data, culture collections, genetic markers, tumor characteristics, regulatory sequences, and more. o Optional introductory course to general metabolism and related computational issues. This will be given prior to the workshop (June 5-7) by Professor Selkov. There will be no scheduled talks. We instead invite open discussion of topics of common interest, such as metabolism and enzymes, or phylogeny, alignments, compounds, motifs, and maps. If you wish to use or generate such integrated biological data, then please join us. Sincerely, Evgeni Selkov Ross Overbeek Natalia Maltsev Amos Bairoch Nat Goodman George Michaels Harold Morowitz Niels Larsen Anthony Bonner Terry Gaasterland Pat Gillevet Rick Stevens Alexander Zamyatnin Sponsorship ----------- The workshop is so far unsponsored, but we welcome interested sponsors. We need help with computer equipment and coverage of travel expenses for Russian participants. Practical details ----------------- Pushchino is a community of some 30,000 people located 100 km south of Moscow. The main acitivity is research. The area is generally safe. The water is not contaminated; the food is fresh and locally produced. You will be able at least to send electronic mail. The registration fee is US $750 for the workshop if you plan to attend the introductory course, $550 otherwise. The fee covers housing (single hotel rooms), meals, and transportation to and from Moscow International Airport. We accept Visa and MasterCard, or a check made out in US$ and cashable within the United States. Upon payment of the workshop fee, participants will be given a US $400 purchase credit for EMP or an actual CD-ROM copy, if there is enough time before the meeting. EMP will soon be generally available for $400 ($4000 for commercial licenses). Registration after the deadline is possible if you pay extra $100 and if there is enough time. Participants need at least tourist visas, which are issued by the nearest Russian consulate, and for which no invitation is needed. To cover your expenses however, your institution may require a business visa, which does require an invitation from the hosts; in that case, please register as soon as possible. Please fill out the fields below; leave blank the fields that do not apply. Your citizenship, date of birth and passport number is needed only if you apply for a business visa. Please send this information, so that we receive it before May 12, by regular mail, fax, or electronic mail, to: Center for Professional Development Business Office, Mail Stop 3G3 George Mason University Fairfax, VA 22030-4444 Fax: 703-993-2112 Electronic mail: overbeek@juju.mcs.anl.gov Questions can be directed to either Ross Overbeek (overbeek@mcs.anl.gov) or Mary Baroody at George Mason University (703-993-2090). ------------------------------------------------------------------- First name: Last name: Research area: Date of birth: Passport number: Citizenship: Tourist visa: Business visa: Institution: Company: Street address: Postal code: City: State: Country: Telephone: Electronic mail: I pay by check: Visa Card #: MasterCard #: Expiration date: Comments: Comments: Comments: Questions: Questions: Questions: ------------------------------------------------------------------- From owner-structural-nmr@net.bio.net Wed Apr 13 23:00:00 1994 Newsgroups: bionet.structural-nmr Path: biosci!bloom-beacon.mit.edu!usc!howland.reston.ans.net!torn!nott!cunews!alfred!mreynold From: mreynold@alfred.carleton.ca (mark reynolds) Subject: CPMAS question Message-ID: Sender: news@cunews.carleton.ca (News Administrator) Organization: Carleton University X-Newsreader: TIN [version 1.2 PL0] Date: Tue, 12 Apr 1994 01:41:18 GMT Lines: 15 I'm currently working on crown ethers derived from diphenylmethane with the crown ring ortho(in both aromatic rings) to the bridging methylene unit. The compound would be 2,2'dimethoxy diphenylmethane to a first approximation. In the uncomplexed form the carbon para to the oxygen(meta to the benzyl carbon) has a chemical shift of ~125ppm in the CP/MAS. Now here's the kicker. Complexed with NaI, that carbon has a shift of ~131ppm!!! This is the only carbon affected by this, so I think I can rule out electronic effects from the cation/anion. Any opinions on what could be the cause of this??? Thanks in advance. Marcus From owner-structural-nmr@net.bio.net Wed Apr 13 23:00:00 1994 Path: biosci!CHEM.UNR.EDU!lcary From: lcary@CHEM.UNR.EDU (lew cary) Newsgroups: bionet.structural-nmr Subject: Re: What is this? Date: 14 Apr 1994 08:48:36 -0700 Organization: BIOSCI International Newsgroups for Molecular Biology Lines: 28 Sender: daemon@net.bio.net Distribution: bionet Message-ID: References: <1994Apr12.093021.8263@uts.uni-c.dk> NNTP-Posting-Host: net.bio.net On Tue, 12 Apr 1994, Mogens Kjaer wrote: > This newsgroup popped up im my newsreader a few days ago, but noone have > posted anything so far - is there something wrong, or are there nobody > out there working with structural NMR? > > Mogens > > -- > Mogens Kjaer > Pronto Software - Development and Distribution > Carlsberg Research Center > Gamle Carlsberg Vej 10 > DK-2500 Valby > Denmark > Phone: + 45 33 27 53 25 > Fax: + 45 33 27 47 08 > Email: carlmk@unidhp.uni-c.dk > > Alright, I'll try. BTW there was an initial post yesterday. I do lots of general nmr, but have Never been at ease interpreting 2D, even COSY. I know that if I studied a 2D plot for a week I would get better, is there an easier way? Most texts (Breitmaier is excellent) do not take you by the hand and gradually increase the complexity. Suggestons? Cheers Lew From owner-structural-nmr@net.bio.net Thu Apr 14 23:00:00 1994 Path: biosci!daresbury!not-for-mail From: Newsgroups: bionet.structural-nmr Subject: PDB file for hum EGF Date: 15 Apr 1994 07:40:18 +0100 Lines: 19 Sender: daemon@mserv1.dl.ac.uk Distribution: bionet Message-ID: <2olcsi$ece@mserv1.dl.ac.uk> Original-To: str-nmr@dl.ac.uk Hi netters, I am trying to find the coordinates (PDB format) for the human EGF. The NMR structure has been published by Cooke et al. Eur. J. Biochem. 193, 807 ff (1990). I could not find anything in the newest release of PDB. OR does anybody know the Email address of Martin BAron or Ian Campbell, working in the Oxford group? thanks Tilman ********************************************************* Tilman Voss Ph.D. Protein Chemistry Group Bender+Co Vienna, Austria Tel: xx43-1-80105 351 Fax: xx43-1-80105 366 Email: VOSS@AIMP.UNA.AC.AT ********************************************************* From owner-structural-nmr@net.bio.net Thu Apr 14 23:00:00 1994 Path: biosci!BIOCH.OX.AC.UK!driscoll From: driscoll@BIOCH.OX.AC.UK (Paul Driscoll) Newsgroups: bionet.structural-nmr Subject: Helium sensor again Date: 15 Apr 1994 06:30:19 -0700 Organization: BIOSCI International Newsgroups for Molecular Biology Lines: 40 Sender: daemon@net.bio.net Distribution: bionet Message-ID: <9404151327.AA18536@bioch.ox.ac.uk> NNTP-Posting-Host: net.bio.net I posted this message locally in USENET but we are not convinced that the bionet.structural-nmr is up an running on this side of the pond yet. Apologies if this is duplicated... It's not really a NMR bio-structural question but... Does anyone out there work in an NMR lab which has an automatic helium sensor or other safety device that is activated in the event of a magnet quench (that leads to some mechanism for automatically venting the helum to the atmosphere, or the evacuation of the building)? These sort of things are probably regarded by most as a bit OTT, but we are having to covince our Health and Safety people that, should a deaf (and blind) person happen to be in the vicinity of the NMR magnet when it quenched, there would be a failsafe mechanism for raising the alarm. I have never seen a catastophic quench. What actually happens? Also, can static magnetic fields (5-10 Gauss) affect in any way the integrity of data-transmisssion lines (telephone or computer network)? It seems highly unlikely to me that there can be any deleterious affect but if anyone has any experience to the contrary (or knows of any recommended specifications for magnet/telephone lines) I would be pleased to hear from you. I am in contact withthe spectrometer and magnet manufacturers, so no need to suggest that - what I am after is real-life experiences. Cheers, Paul Driscoll Royal Society University Research Fellow Department of Biochemistry, University of Oxford, South Parks Road, Oxford OX1 3QU. UK. tel. (44) 865 275799 fax. (44) 865 275259 E-mail: driscoll@bioch.ox.ac.uk From owner-structural-nmr@net.bio.net Fri Apr 15 23:00:00 1994 Path: biosci!bloom-beacon.mit.edu!usc!howland.reston.ans.net!wupost!udel!gandalf.ee.udel.edu!burrill From: burrill@gandalf.ee.udel.edu (Joshua Burrill) Newsgroups: bionet.structural-nmr Subject: Texture Mapping of Neurological MRI Date: 12 Apr 1994 22:58:25 GMT Organization: University of Delaware, Newark Lines: 12 Message-ID: <2of92h$itk@louie.udel.edu> NNTP-Posting-Host: gandalf.ee.udel.edu Hello, I'm involved in a final year project on Texure Mapping of Neurological MRI. I was wondering if anyone had thoguths or had done some work in this field. What I'm particularly interested in is the early detection of Alzhiemer's using texture mapping. So far I've examined the standard first/second order methods using histogram, cross-correlation coefficient etc. Any further ideas would be great. Please Email me, Thanks in Advance, Joshua Burrill (j.burrill@ic.ac.uk) From owner-structural-nmr@net.bio.net Fri Apr 15 23:00:00 1994 Path: biosci!TINMAN.UNL.EDU!rshoe From: rshoe@TINMAN.UNL.EDU (R.Shoe) Newsgroups: bionet.structural-nmr Subject: Re: duplex DNA NMR Date: 16 Apr 1994 12:14:44 -0700 Organization: BIOSCI International Newsgroups for Molecular Biology Lines: 35 Sender: daemon@net.bio.net Distribution: bionet Message-ID: <9404161920.AA27606@tinman.unl.edu> NNTP-Posting-Host: net.bio.net J. Williamson wrote---> >...I believe that he found that glycosidic conformation and pucker were >well determined. However, the values for other torsion angles usually varied >completely over the range of crystallographically observed values. > Of course these torsion angles are often determined by J values from E-Cosy/ DQ-Cosy experiments, and as Harbison has recently pointed out (which was further confirmed by Drobney in an excellent poster at ENC) these values can be severely scaled by differential cross-relaxation effects (I should say the directly measured values). >You didn't mention if you had used relaxation matrix methods or not. They are reported >to help some. I personally have no experience with those methods because I bagged it >before they came around. > >The problem is that nucleic acids, in particular DNA, have a very different topology for >the spins compared to proteins. There are very few interresidue NOEs that connect nucleotides >that are not adjacent. We get a very local picture of what is going on. > >Anyone care to vouch for the efficacy of relaxation matrix stuff? > > I am fairly familiar with the rel.matrix methods popularized by James, et.al, and I think that there is little doubt that they increase the reliability of calculated NOE distances. By taking into account the relaxation effects on the observed J-splitting (as mentioned above), torsion angles can be nailed down with fairly decent accuracy as well. So, although difficult, I don't think that NMR structure studies of nucleotides is THAT hopless (IMHO). Best Regards, Richard Shoemaker Chemistry, Univ. of Nebraska-Lincoln. From owner-structural-nmr@net.bio.net Fri Apr 15 23:00:00 1994 Path: biosci!PICASSO.UCSF.EDU!schmitz From: schmitz@PICASSO.UCSF.EDU Newsgroups: bionet.structural-nmr Subject: Duplex DNA NMR Date: 16 Apr 1994 13:47:04 -0700 Organization: BIOSCI International Newsgroups for Molecular Biology Lines: 55 Sender: daemon@net.bio.net Distribution: bionet Message-ID: <199404162047.NAA15878@dali.ucsf.edu> NNTP-Posting-Host: net.bio.net I just came back from the ENC meeting, where they had a lot of very interesting stuff: among pulse sequences, simulation programs and lots of applications, there were not to many posters on DNA, especially not simple double-helical DNA. We all know that the focus in NMR and nucleic acids has shifted to RNA and more unusual DNA molecules and that has a lot to do with the frustration that is reflected in several papers and the initial comment of David Iain Greig. I consider myself one of those converts. But nevertheles, I think, it is not at all that grim and there has been a gradual progress in the field. Here are some of my experiences and - opinions: I have done two DNA octamer structures with restrained MD using, in one case NOE restraints alone, and in another NOE plus torsion angle restraints (and also a lot of MD stuff). Here, in the Tom James group, we use a complete relaxation matrix derived distances - which increases the accuracy of the restraints and also - you get more of them. To me, the number of restraints is very crucial. If there were enough you didn't even need a force-field to do the rest of the job - another tricky issue. I agree with Greig and Dave Wemmer when I look at my first structure and other "earlier" DNA work. But things look different for a more recent example, where each nucleotide had 20 distance restraints on average not counting torsion angle pucker restraints. In that case, restraints could even be thrown out and the rMD structure did not change (free R-factor analysis, Weisz et al. BIOCHEMISTRY, 1994, 33, 354-366). The improvement came mainly from iterative peak deconvolution programs, incorporation of imino distances (requiring measurement of exchange rates and linear excitation profiles) and pucker restraints. There was really little slack in the structure and one can be more comfortable wit the structural parameters obtained. Of course, some of the structural parameters are more defined than other ones and it still seems tricky to elucidate the "true, sequence dependent parameters". One way of getter a fairer picture is to get some error bounds on the helical parameters, e.g. one could run a longer rMD trajectorie with moderate forceconstants and analyze the distribution of the struc.parameters. Some things will be completely washed out . For example,we got some striking similarities for TpG steps in three different structures. My biggest concern, however, has to do with the dynamic effects that we know so little about. NMR structures make sense only if we are dealing with a clear major conformer and in some cases we might not have that according to coupling constants analysis. Also here, some interesting posters at the ENC (Reid, Griesinger, Drobny): with longer correlation times, coupling constants are not what they appear. However, for small fragments or corr. times below 5ns, the effect seemed negligible regarding the experimental error bounds on coupl. cons. In my view, the problem spot is shifting from the old underdefinedness problems to more methodological problems, i.e.averaging artefacts. I would be interested in hearing other folks opinion on that. Another point that I want to raise is the very sensitivity of rMD protocols (MD is a fleebag of trouble). E.g., there are a lot of ways to produce bend DNA with AMBER, but there are also ways to prevent some of these artefacts. Bach to David Greig's example: >I do see certain patterns in the NOE spectrum, such as very weak (1',8) >sequential peaks in the second T-A step in *both* strands, indicating either >a kink or propellor twist or the like, but the RMS deviation in my rMD/SA >structures (using X-PLOR 3.1) is over an angstrom, even if I do get >minimal violations of my NOE constraints. And the helix is underwound. >Sigh. These anomalities are very exciting, especially if there are other things in the NMR data that are correlated. There seems to be something going on with certain TpA steps: I had a similar situation in my second structure (GTATAATG)(CATATTAC), where the A5 protons had much broader lines, esp. H2 (JMB,1992 V227 510-531). Also B.Reid's group had a similar experience with a TpA step (Biochemistry 93, 32, 8022-8035). These things, I think, we can still be very excited about as NMR is pretty much the only tool to learn more about them. (I apologize for the self-indulgingly long message - but it's been about the pain and the joy of my last five years...) Uli ><><><><><><><><><><><><><><><><><><> Uli Schmitz UC,San Francisco Dept. Pharmaceutical Chemistry Phone (415) 476 4378; Fax (415) 476 0688; schmitz@picasso.ucsf.edu ><><><><><><><><><><><><><><><><><><> From owner-structural-nmr@net.bio.net Sat Apr 16 23:00:00 1994 Newsgroups: bionet.structural-nmr Path: biosci!CS.Arizona.EDU!uunet!solaris.cc.vt.edu!news.ans.net!howland.reston.ans.net!news.intercon.com!psinntp!psinntp!psinntp!newsserver.pixel.kodak.com!rpi!utcsri!utnut!nott!cunews!alfred!mreynold From: mreynold@alfred.carleton.ca (mark reynolds) Subject: CP/MAS Message-ID: Sender: news@cunews.carleton.ca (News Administrator) Organization: Carleton University X-Newsreader: TIN [version 1.2 PL0] Date: Tue, 12 Apr 1994 19:29:05 GMT Lines: 13 Let me set the stage. I have synthesized a crown ether derived from diphenylmethane. A first approximation of this molecule would be 2,2'dimethoxy diphenylmethane, or bis(2methoxyphenyl)methane. In the CP/MAS the carbon para to the methoxy unit(meta to the benzyl carbon) has a shift of approximately 120ppm. Now here's the kicker, when complexed with NaI, the chemical shift of that carbon is now ~131ppm!! Since only that carbon changes significantly, I believe it is not a electronic effect but a stereochemical effect. Anybody have an idea what could cause such a significan shift??? Thanks in advance Marcus From owner-structural-nmr@net.bio.net Sun Apr 17 23:00:00 1994 Path: biosci!PRIMER.MED.YALE.EDU!barry From: barry@PRIMER.MED.YALE.EDU (Barry Schweitzer) Newsgroups: bionet.structural-nmr Subject: 13-C labelled cholesterol Date: 18 Apr 1994 13:24:56 -0700 Organization: BIOSCI International Newsgroups for Molecular Biology Lines: 22 Sender: daemon@net.bio.net Distribution: bionet Message-ID: <9404182025.AA02428@primer.med.yale.edu> NNTP-Posting-Host: net.bio.net Hello! Does anyone out there know where I can get my hands on some 13-C labelled cholesterol? Or has anyone seen any papers in which labelled cholesterol has been used? Thanks in advance. Barry ========================================================================== Barry Schweitzer, Ph.D. Department of Pediatrics "Whoever loves discipline, Yale University School of Medicine loves knowledge New Haven, CT 06510 But he who hates correction 203-785-6780 Fax: 203-785-7194 is stupid." Proverbs 12:1 email: barry@primer.med.yale.edu ========================================================================== From owner-structural-nmr@net.bio.net Sun Apr 17 23:00:00 1994 Path: biosci!SCRIPPS.EDU!jmount From: jmount@SCRIPPS.EDU (John Mountzouris) Newsgroups: bionet.structural-nmr Subject: (none) Date: 18 Apr 1994 09:24:26 -0700 Organization: BIOSCI International Newsgroups for Molecular Biology Lines: 15 Sender: daemon@net.bio.net Distribution: bionet Message-ID: <9404181624.AA05149@wright.Scripps.EDU> NNTP-Posting-Host: net.bio.net April 18, 1994 To structural NMR network: I am a postdoctoral student in Peter Wright's laboratory at Scripps studying protein structure. I recently completed doctoral work with Laurence Hurley at The University of Texas studying drug-nucleic acid complex structure. I would like to know of the network has been formed, and if so, I would like to participate. Thank you. Truly, John Mountzouris email: jmount@wright.scripps.edu From owner-structural-nmr@net.bio.net Sun Apr 17 23:00:00 1994 Path: biosci!CHEM.PURDUE.EDU!David From: David@CHEM.PURDUE.EDU Newsgroups: bionet.structural-nmr Subject: Re: duplex DNA NMR Date: 18 Apr 1994 12:30:30 -0700 Organization: BIOSCI International Newsgroups for Molecular Biology Lines: 43 Sender: daemon@net.bio.net Distribution: bionet Message-ID: <199404181930.MAA15775@net.bio.net> NNTP-Posting-Host: net.bio.net David Iain Greig writes: >I am currently finishing up my M.Sc. in Biophysics at the University of Toronto >working on the NMR-derived structure of a tetradecameric DNA duplex >d(GGGCTATAAAAGGG)-d(CCCTTTTATAGCCC). In writing up my thesis, I am forced >to conclude that 1H NMR of ordinary B-form DNA is of limited utility in >discerning fine structure present in the duplex. >I do see certain patterns in the NOE spectrum, such as very weak (1',8) >sequential peaks in the second T-A step in *both* strands, indicating either >a kink or propellor twist or the like, but the RMS deviation in my rMD/SA >structures (using X-PLOR 3.1) is over an angstrom, even if I do get >minimal violations of my NOE constraints. And the helix is underwound. Dear Iain, I want to confirm the message from others about the importance of a large number of high quality contraints for solving the structures of duplex DNA. In Kaluarachichi et al. (Biochemistry, 30, 8785 (1991), Kumar K., Rob Meadows and I carried out a series of simulations to test the accuracy and precision of structures, comparing two-spin vs. relaxation Matrix methods. The bottom line is that if you use enough of the constraints with matrix methods you can reproduce the local helical parameters pretty well. Even torsion angles can be reproduced even though no torsional restraints were included. I should add that our MORASS program for doing a hybrid relaxation matrix method analysis is available over internet using anonymous ftp or Gopher. Good luck! __________________________________________________________________ Dr. David Gorenstein Chemistry Department Purdue University W. Lafayette, IN 47907 USA Internet: David@chem.purdue.edu Office: (317) 494 7851 FAX: (317) 494 0239 From owner-structural-nmr@net.bio.net Sun Apr 17 23:00:00 1994 Path: biosci!PICASSO.UCSF.EDU!ulyanov From: ulyanov@PICASSO.UCSF.EDU (Nikolai Ulyanov) Newsgroups: bionet.structural-nmr Subject: Re: Welcome, welcome... Date: 18 Apr 1994 16:38:51 -0700 Organization: BIOSCI International Newsgroups for Molecular Biology Lines: 59 Sender: daemon@net.bio.net Distribution: bionet Message-ID: <199404182339.QAA29494@picasso.ucsf.EDU> NNTP-Posting-Host: net.bio.net David Iain Greig writes: >...... I am forced >to conclude that 1H NMR of ordinary B-form DNA is of limited utility in >discerning fine structure present in the duplex. Dave, it's hard to argue with your statement.. Indeed, it's much easier to get useful information for molecules with unknown overall structure -- any tertiary contact is very informative, even if the distance is estimated very roughly. For DNA duplexes, the answer is known in advance -- it is B-form; if you want to learn something else, you need to take pains of increasing the number and accuracy of your restraints. Actually, previous posts gave the lines along which this can be done: - relaxation matrix methods, they are really better than ISPA; there are a number of programs available -- to calculate distances accounting for spin diffusion, or to do refinement directly against intensities. - how to get more *important* restraints -- there are just a few inter-strand distances which can be measured (if you are lucky) in the DNA duplex, but they can make a difference. Distances involving exchangeable protons can be very helpful, especially if you actually measured the exhange rates. Another type of inter-strand distances is Ade H2 - H1', there must be some in your duplex with A-tract. In the alternating TATA part, look carefully at inter- strand Ade H2 - Ade H2 contacts. My experience shows that if you don't inter-strand distances, helical parameters are defined very poorly, even if you have plenty of other distances (1992 Biochemistry 31, 3918). - I don't remember if anybody mentioned so called "non-NOE". E.g., if the distance Ade H2 - H1' across the groove is long (so that you don't see NOE), it is also very important information. Still, you are not guaranteed to get an accurate structure. Ideally, one needs some independent *solution* data. There are not much of them around, but there are some. E.g., average helical twist was measured very reliably. > ... And the helix is underwound. >Sigh. How much underwound? Average hel.twist in solution is 34.5 deg. (compared to 36.0 deg. in crystal). If it is underwound much more than that, something is wrong. One thing to look at would be the energy term. In some earlier studies on Dickerson dodecamer (cited, e.g., by Wemmer), energy was completely neglected, the idea was to use experimental data only. It didn't work: the DNA was severely underwound. I know that X-plor lets you scale down the energy during the simulated annealing -- make sure that you scale it back and run MD sufficiently long with full energy term after the cooling. Nick ----------------------------------------------------------- Nikolai B. Ulyanov Dept. Pharmaceutical Chemistry UCSF San Francisco, CA 94143-0446 ulyanov@picasso.ucsf.edu ------------------------------------------------------------ From owner-structural-nmr@net.bio.net Mon Apr 18 23:00:00 1994 Path: biosci!CS.Arizona.EDU!noao!asuvax!cs.utexas.edu!howland.reston.ans.net!wupost!news.miami.edu!usenet.ufl.edu!Micro!marian From: marian@Micro.uucp (Marian Buszko) Newsgroups: alt.sci.nmr,bionet.structural-nmr Subject: X-tracts Date: 18 Apr 1994 19:55:31 GMT Organization: University of Florida Lines: 6 Distribution: na Message-ID: <2ouojjINNm1i@no-names.nerdc.ufl.edu> NNTP-Posting-Host: micro.ifas.ufl.edu Marian === Marian Buszko (marian@micro.ifas.ufl.edu) Univ. of Florida, Microbiology From owner-structural-nmr@net.bio.net Mon Apr 18 23:00:00 1994 Path: biosci!DINGO.NIEHS.NIH.GOV!tom From: tom@DINGO.NIEHS.NIH.GOV (Thomas M. O'Connell) Newsgroups: bionet.structural-nmr Subject: (none) Date: 19 Apr 1994 06:36:42 -0700 Organization: BIOSCI International Newsgroups for Molecular Biology Lines: 19 Sender: daemon@net.bio.net Distribution: bionet Message-ID: <9404191337.AA04771@dingo.niehs.nih.gov> NNTP-Posting-Host: net.bio.net Dear Diana, Several members of my group have been involved in the studies using PCA extracts for subsequent 31P NMR analysis. A helpful reference would be "Uridine Diphospho Sugars and Related Hexose Phosphates in the Liver of Hexosamine Treated Rats: Identification Using 31P-1H 2D NMR with HOHAHA Relay", Perlman et. al. Biochem. 1990, 29, 4318. Hope this helps. Tom ******************************************************* Thomas M. O'Connell Ph.D. Laboratory of Molecular Biophysics National Institute of Environmental Health Sciences tom@dingo.niehs.nih.gov ph 919-541-3872 fax -7880 From owner-structural-nmr@net.bio.net Mon Apr 18 23:00:00 1994 Path: biosci!PRIMER.MED.YALE.EDU!barry From: barry@PRIMER.MED.YALE.EDU (Barry Schweitzer) Newsgroups: bionet.structural-nmr Subject: Re: DNA duplex structures by NMR Date: 19 Apr 1994 08:22:13 -0700 Organization: BIOSCI International Newsgroups for Molecular Biology Lines: 63 Sender: daemon@net.bio.net Distribution: bionet Message-ID: <9404191522.AA03119@primer.med.yale.edu> NNTP-Posting-Host: net.bio.net Hello! Well, the first posting to this newsgroup was obviously a good one in that it has sparked alot of good follow-ups (Thanks, David!). Before these follow-ups peter out, I'd like to get my two cents in. I believe that more recent postings have been more on the mark: it is possible to get informative structures of DNA duplexes by NMR, but it is not easy! (Does anyone know of any type of structure that is easy to obtain by NMR? If so, please tell me so I can switch my focus!). It has been stated that the number and quality of the restraints is also critical. The quality of the restraints (i.e. the absolute need for relaxation matrix analysis) and the need for as many inter-residue and inter-strand restraints has already been covered. I submit that using as many backbone restraints as possible may be just as important. Some say that it is the backbone that determines the structure of DNA and the bases just go along for the ride. I wouldn't go that far, but too often, DNA structures are calculated without any constraints from the H3', H4', H5'/H5", or phosphorus. Yes, they are harder to assign due to overlap and spin diffusion is obviously a problem, but experiments are out there that make the assignment easier, and relaxation matrix analysis helps out with the spin diffusion. X-PLOR also deals fairly effectively with overlapping signals. I also advocate the type of analysis described by Kim, Lin & Reid (Biochemistry (1992) 31, 3564) and others for determining sugar pucker. This method uses a large number of *qualitative* pieces of data to constrain the sugar pucker to a relatively narrow range. Finally, I'd like to briefly revisit what several members of the James lab have said in this forum as well as in their recent papers in Biochemistry, J.Mol.Bio., etc. My first point concerns the rMD protocol one uses. I have taken my cue from procedures used by the James lab in which experimental restraints are heavily weighted during the heating and high temperature phases, while the force field constraints are kept low. This priority order is reversed during cooling and equilibration. I have found that this gives much "tighter" structures than more uniform protocols. It also appears useful for pointing out restraints that may be "wrong". (David, I believe you also use X-PLOR, so I would be happy to send you my script if you are interested.) My second point is that the James group is probably right on the mark when they say that the real story in DNA may lie in the dynamics of the molecule. Unfortunately, current refinement procedures do little, if anything, to take the time averaged nature of the restraints into account, although the James group is trying. Recent papers have appeared that examine the dynamics of the sugar, and my lab is using some new experiments to look at the dynamics of the phosphate group, so maybe there is reason to be hopeful on this front. Sorry for the long posting, but like some of the others, this issue has occupied my thoughts for a long time. Barry ========================================================================== Barry Schweitzer, Ph.D. Department of Pediatrics "Whoever loves discipline, Yale University School of Medicine loves knowledge New Haven, CT 06510 But he who hates correction 203-785-6780 Fax: 203-785-7194 is stupid." Proverbs 12:1 email: barry@primer.med.yale.edu ========================================================================== From owner-structural-nmr@net.bio.net Mon Apr 18 23:00:00 1994 Path: biosci!headwall.Stanford.EDU!unixhub!fnnews.fnal.gov!nntp-server.caltech.edu!netline-fddi.jpl.nasa.gov!elroy.jpl.nasa.gov!news.msfc.nasa.gov!sol.ctr.columbia.edu!howland.reston.ans.net!wupost!news.miami.edu!usenet.ufl.edu!dgarside From: dgarside@ufnmr1.cis.ufl.edu (Diana Garside) Newsgroups: bionet.structural-nmr,alt.sci.nmr Subject: extracts Date: 18 Apr 1994 20:05:35 GMT Organization: Univ. of Florida JHMHC Dept. of Radiology Lines: 7 Distribution: world Message-ID: <2oup6fINNm6l@no-names.nerdc.ufl.edu> NNTP-Posting-Host: ufnmr1.health.ufl.edu Does anyone have any helpful hints how to successfully do perchloric acid extracts for 31P NMR studies? Diana dgarside@ufnmr1.health.ufl.edu From owner-structural-nmr@net.bio.net Mon Apr 18 23:00:00 1994 Path: biosci!CS.Arizona.EDU!noao!asuvax!cs.utexas.edu!howland.reston.ans.net!wupost!news.miami.edu!usenet.ufl.edu!Micro!marian From: marian@Micro.uucp (Marian Buszko) Newsgroups: alt.sci.nmr,bionet.structural-nmr Subject: X-tracts Date: 18 Apr 1994 19:51:29 GMT Organization: University of Florida Lines: 1 Message-ID: <2ouoc1INNm10@no-names.nerdc.ufl.edu> NNTP-Posting-Host: micro.ifas.ufl.edu From owner-structural-nmr@net.bio.net Mon Apr 18 23:00:00 1994 Path: biosci!UMDIX.UMDC.UMU.SE!peterl From: peterl@UMDIX.UMDC.UMU.SE (Peter Lundberg Fysikalisk Kemi) Newsgroups: bionet.structural-nmr Subject: Re: extracts Date: 19 Apr 1994 00:51:28 -0700 Organization: BIOSCI International Newsgroups for Molecular Biology Lines: 1 Sender: daemon@net.bio.net Distribution: bionet Message-ID: <199404190751.AA25764@umdix.umdc.umu.se> NNTP-Posting-Host: net.bio.net From owner-structural-nmr@net.bio.net Tue Apr 19 23:00:00 1994 Path: biosci!CS.Arizona.EDU!uunet!pipex!warwick!lyra.csx.cam.ac.uk!mole.bio.cam.ac.uk!arcr1 From: arcr1@mole.bio.cam.ac.uk (Andy Raine (Bioc)) Newsgroups: bionet.structural-nmr Subject: Re: DNA duplex structures by NMR Date: 20 Apr 1994 08:15:29 GMT Organization: U. of Cambridge, England Lines: 34 Distribution: bionet Message-ID: <2p2ob1$n2t@lyra.csx.cam.ac.uk> References: <9404191522.AA03119@primer.med.yale.edu> NNTP-Posting-Host: mole.bio.cam.ac.uk In article <9404191522.AA03119@primer.med.yale.edu>, barry@PRIMER.MED.YALE.EDU (Barry Schweitzer) writes: |> Finally, I'd like to briefly revisit what several members of the James lab |> have said in this forum as well as in their recent papers in Biochemistry, |> J.Mol.Bio., etc. My first point concerns the rMD protocol one uses. I have |> taken my cue from procedures used by the James lab in which experimental |> restraints are heavily weighted during the heating and high temperature |> phases, while the force field constraints are kept low. This priority order |> is reversed during cooling and equilibration. I have found that this |> gives much "tighter" structures than more uniform protocols. It also appears |> useful for pointing out restraints that may be "wrong". I may be missing something here, but this worries me a little. Do structures calculated using the "James weighting" agree with your experimental data better or worse than those calculated with a more conventional protocol? If the fit is no better, then an ensemble with a greater RMSD would be a more honest interpretation of your data. If the structures fit the data better, as well as giving a lower RMSD then there is no problem of course! Andrew -------------------------------------------------------------------------------- Dr. Andrew Raine +44 223 333744 Department of Biochemistry or 333499 University of Cambridge Tennis Court Road Cambridge CB2 1QW United Kingdom -------------------------------------------------------------------------------- From owner-structural-nmr@net.bio.net Tue Apr 19 23:00:00 1994 Path: biosci!NET.BIO.NET!biosci-help From: biosci-help@NET.BIO.NET (BIOSCI Administrator) Newsgroups: bionet.structural-nmr Subject: IMPORTANT - Mailing list is being split Date: 20 Apr 1994 12:34:19 -0700 Organization: BIOSCI International Newsgroups for Molecular Biology Lines: 88 Sender: daemon@net.bio.net Distribution: bionet Message-ID: Reply-To: biosci-help@net.bio.net NNTP-Posting-Host: net.bio.net Now that sufficient time has elapsed to allow people to unsubscribe if they wish to use USENET news instead of e-mail, I have sent the European, African, and Central Asian addresses below to our U.K. node in Daresbury. As soon as I receive confirmation that they have been installed there I will delete them from the list here. It is possible that those of you below might receive duplicate copies of STR-NMR mail for a short period during this changeover. If your address is included in the list below, please use the biosci@daresbury.ac.uk address in the future for any mail regarding your e-mail subscriptions. Please DO NOT post subscribe/unsubscribe requests to the newsgroup str-nmr posting addresses!!!! The newsgroup posting address at Daresbury is str-nmr@daresbury.ac.uk and should be used by the people below for posting messages to the entire readership. Sincerely, Dave Kristofferson BIOSCI/bionet Manager biosci-help@net.bio.net arcr1@mbuf.bio.cam.ac.uk berger@ps1515.chemie.uni-marburg.de bgc0829@ggr.co.uk billeter@mol.biol.ethz.ch bolis@icil64.cilea.it breeze%iciukph.uucp@eros.britain.eu.net breg@ruuci9.chem.ruu.nl bryan@freja.fkem2.lth.se chary@tifrvax.tifr.res.in debbie@chemsg5.tau.ac.il dietmar@akjung3.orgchemie.chemie.uni-tuebingen.de domel@alfa.ichf.edu.pl drablos@marvin.mr.sintef.no driscoll@bioch.ox.ac.uk edith@ruuci2.chem.ruu.nl erard@ibcg.biotoul.fr franky.fant@rug.ac.be frits@rulglj.leidenuniv.nl g.a.morris@man.ac.uk gharden@phoenix.oulu.fi gkb1@york.ac.uk guentert@mol.biol.ethz.ch h475nes@ella.hu hong@indra.chem.umu.se ikilpela@introni.helsinki.fi imtjs26@cc.csic.es iourine@lourie.und.ac.za jackson@psipsy.uct.ac.za janssens@agr.kuleuven.ac.be jap01@castle.ed.ac.uk jmw@biochemistry.oxford.ac.uk jose@orgc2.vub.ac.be jrene@tome.cbs.univ-montp1.fr kalle@ruuci9.chem.ruu.nl kris.boulez@rug.ac.be kris@tome.cbs.univ-montp1.fr kristian@tome.cbs.univ-montp1.fr mbarg@dl.ac.uk mbdpsgm@cms.mcc.ac.uk mrigank@imtech.ernet.in mzloh@ic.ac.uk nmrlist@tome.cbs.univ-montp1.fr nmueller@jk.uni-linz.ac.at peterl@umdix.umdc.umu.se pjk@oyster.csb.ki.se placebio-dali@macpost.lu.se rasmus@ruuci11.chem.ruu.nl rgordon@ucmb.ulb.ac.be rolf@nmr.varian.ch ronald@nmr.chem.ruu.nl roylanc@max.mpibp.uni-frankfurt.de rull@nmr.chem.ruu.nl rwoods@biop.ox.ac.uk smb18@mbfs.bio.cam.ac.uk smvf100@cus.cam.ac.uk stote@borneo.u-strasbg.fr terez@tome.cbs.univ-montp1.fr thep@risc1.lrm.fi.cnr.it uccaaea@ucl.ac.uk vberckm@schs.uia.ac.be wgalazka@chem.uw.edu.pl wim.vranken@rug.ac.be From owner-structural-nmr@net.bio.net Tue Apr 19 23:00:00 1994 Path: biosci!rutgers!uwvax!uchinews!vixen.cso.uiuc.edu!news.uoregon.edu!netnews.nwnet.net!news.u.washington.edu!kantola From: kantola@u.washington.edu (Angeline Kantola) Newsgroups: bionet.structural-nmr Subject: Re: What is this? Message-ID: <2oi17u$b1n@news.u.washington.edu> Date: 14 Apr 94 00:03:10 GMT References: <1994Apr12.093021.8263@uts.uni-c.dk> Organization: University of Washington, Seattle Lines: 18 NNTP-Posting-Host: carson.u.washington.edu In article <1994Apr12.093021.8263@uts.uni-c.dk>, Mogens Kjaer wrote: >This newsgroup popped up im my newsreader a few days ago, but noone have >posted anything so far - is there something wrong, or are there nobody >out there working with structural NMR? Sure there are! OK, I've got something to ask. I'm a graduate student in Biochemistry, studying protein structure via (surprise) NMR. Recently I found out about and poked around in the NMRFAM gopher site out of the University of Wisconsin; a variety of NMR-related programs are accessible through this site. (Maybe someone who's affiliated with the NMR center at the other UW could fill us in a little more about NRMFAM?) Aside from the PDB site itself, are there other gopher or ftp sites related to protein structure or strucutural NMR? --Angie From owner-structural-nmr@net.bio.net Tue Apr 19 23:00:00 1994 Path: biosci!PICASSO.UCSF.EDU!schmitz From: schmitz@PICASSO.UCSF.EDU Newsgroups: bionet.structural-nmr Subject: Re: DNA duplex structures by NMR Date: 20 Apr 1994 11:18:00 -0700 Organization: BIOSCI International Newsgroups for Molecular Biology Lines: 31 Sender: daemon@net.bio.net Distribution: bionet Message-ID: <199404201818.LAA17747@dali.ucsf.edu> NNTP-Posting-Host: net.bio.net Hello again, I would like to respond to the issue raised by Andrew Raine and Barry Schweitzer; Andrew wrote: >I may be missing something here, but this worries me a little. Do structures >calculated using the "James weighting" agree with your experimental data better>or >worse than those calculated with a more conventional protocol? If the fit is n>o better, then an ensemble with a greater RMSD would be a more honest interpret>ation of your data. >If the structures fit the data better, as well as giving a lower RMSD then there >is no problem of course! -------------------------- Our simulated annealing protocols were prompted by problems when starting structures were used that are far away from the target, e.g. A- or Z-DNA for a B-ish solution structure. In the beginning I personaly had a lot of respect for MM and MD programs. Now, I would say, you could anything to your model in the early stage of your rMD refinement as long you give enough time to equilibrate with the full force field and moderate restraint weighting. Most structural parameters will depend to some extend on the ratio of force field and restraint contribution! I have heard that with XPLOR you can even fold a single strand RNA piece into a hairpin by tweeking the force field. But whatever tightly converged structures you get in your rMD runs with different starting structures, the rmsd is only about the precision and unfortunately not about the accuracy of the structure. But high precision means that you have reached roughly the same energy minimum from different starting points, important but not everything. Even going for the best match with the experimental data is not trivial, because you can always force a better fit with higher force constants - overfitting is really easy. And the balance is slightly different for each molecule it seemed. I think it is helpful to get an idea about the relationship between the force constants and the energies (constraint energy and total conformational energy). A particular problem with overfitting involves the susceptability of most MD programs to bending and twisting of HC-bonds, baseplains and more - but that is where often better R-values and low rmsd values stem from. I guess, this problem has not been addressed too explicitly in the literature. However, verification of the accuracy of our NMR structures is still the big challenge. Uli ><><><><><><><><><><><><><><><><><><> Uli Schmitz UC,San Francisco Dept. Pharmaceutical Chemistry Phone (415) 476 4378; Fax (415) 476 0688; schmitz@picasso.ucsf.edu ><><><><><><><><><><><><><><><><><><> From owner-structural-nmr@net.bio.net Tue Apr 19 23:00:00 1994 Path: biosci!rutgers!utcsri!utnut!oci!greig From: greig@tris.oci.utoronto.ca (David Iain Greig) Newsgroups: bionet.structural-nmr Subject: Welcome II Keywords: NMR DNA MARDIGRAS rMD/SA Message-ID: Date: 19 Apr 94 16:21:22 GMT Sender: greig@tris (David Iain Greig) Followup-To: bionet.structural-nmr Organization: Structural Biology, Ontario Cancer Institute Lines: 58 Ok... I should have been more explicit... =^P In my structural calculations, I am using rMD/SA using X-PLOR 3.1, with a modified version of the sa.inp protocol, starting from a B-DNA model generated by Sybyl. sequence : GGGCTATAAAAGGG CCCGATATTTTCCC I am using a personally-modified version of parallhdg.dna/topallhdg.dna, which I have mentioned previously on the x-plor mailing list. Basically, I changed some bond angles to conform to Saenger's values in the sugar, and some methyl nomenclature. I obtain peak volumes by manual integration using NMRz from a 150ms NOESY spectrum, and have checked them by peakfitting using NMRz as well for theoretical peakshapes: the values agree fairly well. I then use MARDIGRAS to solve the relaxation matrix; the problem is that due to *serious* overlap in my 2D spectrum, I can't get 'resolved' peak volumes for over half my crosspeaks, and have to enter them as 'unresolved' values in MARDIGRAS. I *do* get reasonable distance estimates from MARIDGRAS, but I find it has problems with (1'/3', Ar) and (1'/3',Me) peaks, presumably due to *heavy* spin diffusion from the 2'/2" pair. But allowing an increased upper bound is a simple solution for this.... The 'unresolved' peaks are used as strong/medium/weak restraints in the MD simulation. I also include explicit hbond distance restraints to account for Watson-Crick basepairing. (Plus labile restraints from a solvent- supression NOESY spectrum...) I have tried the Clore&Gronenborn 'theoretical' backbone torsion restraints as the PECOSY we ran on the sample was also severely overlapped, and we really didn't want to try modelling the crosspeaks to extract coupling constants. I don't like using them, because they're non-experimental. Anyone else have any comments on them? Summary: Relaxation matricies are probably the best way to get enough information to produce an accurate structure, IMHO. The alternative, of measuring peak volumes at very short mixing times, a la Clore&Gronenborn, seems to have some theoretical problems, I gather. The problem arises in the overlap in my sequence: two runs of GGG and a run of AAAA. Bleah. At least I get good dispersion in the TATA region! But I am in agreement that *some* interesting information can be gathered. I just don't think the 'state of the art' is sufficiently advanced to give the old '0.2 A RMS' structures that protein NMR types have come to demand. =^) Thanks for the responses! --D. -------------------------------------------------------------------------- David Iain Greig greig@oci.utoronto.ca Dept. of Medical Biophysics RIPEM/PGP Public Keys available on request University of Toronto "A clean bathroom makes a quiet mind" Ontario Cancer Institute/Princess Margaret Hospital Kibo Number 1 -------------------------------------------------------------------------- From owner-structural-nmr@net.bio.net Tue Apr 19 23:00:00 1994 Path: biosci!PRIMER.MED.YALE.EDU!barry From: barry@PRIMER.MED.YALE.EDU (Barry Schweitzer) Newsgroups: bionet.structural-nmr Subject: Re: DNA duplex structures by NMR Date: 20 Apr 1994 06:53:21 -0700 Organization: BIOSCI International Newsgroups for Molecular Biology Lines: 69 Sender: daemon@net.bio.net Distribution: bionet Message-ID: <9404201354.AA03872@primer.med.yale.edu> NNTP-Posting-Host: net.bio.net > > In article <9404191522.AA03119@primer.med.yale.edu>, barry@PRIMER.MED.YALE.EDU (Barry Schweitzer) writes: > > > > |> Finally, I'd like to briefly revisit what several members of the James lab > |> have said in this forum as well as in their recent papers in Biochemistry, > |> J.Mol.Bio., etc. My first point concerns the rMD protocol one uses. I have > |> taken my cue from procedures used by the James lab in which experimental > |> restraints are heavily weighted during the heating and high temperature > |> phases, while the force field constraints are kept low. This priority order > |> is reversed during cooling and equilibration. I have found that this > |> gives much "tighter" structures than more uniform protocols. It also appears > |> useful for pointing out restraints that may be "wrong". > > I may be missing something here, but this worries me a little. Do structures > calculated using the "James weighting" agree with your experimental data better or > worse than those calculated with a more conventional protocol? If the fit is no > better, then an ensemble with a greater RMSD would be a more honest interpretation > of your data. > > If the structures fit the data better, as well as giving a lower RMSD then there > is no problem of course! > > Andrew > > -------------------------------------------------------------------------------- > Dr. Andrew Raine +44 223 333744 > Department of Biochemistry or 333499 > University of Cambridge > Tennis Court Road > Cambridge > CB2 1QW > United Kingdom > -------------------------------------------------------------------------------- > Thanks for pointing this out! I see now that this part of my posting was incomplete. Using protocols adapted from the James lab gave me a family of structures with somewhat lower RMSD than protocols in which the weighting of the experimental data was kept uniform (for my duplex, 0.7 A vs 0.9 A). So "much 'tighter'" was not a correct description of my results. There was also better agreement between structures calculated from a canonical B-type starting structure and those calculated from a canonical A-type structure, indicating that this protocol appears to do a better job of sampling conformational space. These structures also gave lower R(1/6) factors, indicating a better fit with the experimental data. I also should not have implied that the weighting was the only factor in this. The James protocol uses a slow heating, a few psec of high temperature, followed by cooling to a final equilibrium temperature (300K). My earlier protocols used two phases, first at a moderately high temperature, then at 300K for equilibration but both using the same weighting schemes. This difference in temperature control undoubtedly led to a better sampling of conformational space. What I need to do is run my new protocol with constant weighting and see what happens. I'll post the results of this. ========================================================================== Barry Schweitzer, Ph.D. Department of Pediatrics "Whoever loves discipline, Yale University School of Medicine loves knowledge New Haven, CT 06510 But he who hates correction 203-785-6780 Fax: 203-785-7194 is stupid." Proverbs 12:1 email: barry@primer.med.yale.edu ========================================================================== From owner-structural-nmr@net.bio.net Wed Apr 20 23:00:00 1994 Path: biosci!MICRO.IFAS.UFL.EDU!marian From: marian@MICRO.IFAS.UFL.EDU (Marian Buszko) Newsgroups: bionet.structural-nmr Subject: Re: What is this? Date: 21 Apr 1994 06:22:17 -0700 Organization: BIOSCI International Newsgroups for Molecular Biology Lines: 31 Sender: daemon@net.bio.net Distribution: bionet Message-ID: References: Reply-To: Marian Buszko NNTP-Posting-Host: net.bio.net On 20 Apr 1994, Lawrence McIntosh wrote: > A few useful gopher sites are: (stuff deleted) > National Magnetic Resonance Facility at Madison (NMRFAM) > gopher://micro.ifas.ufl.edu/11/NMR_Information_Server Fri Mar 18 12:10:47 1994 > NMR Information Server Great list, Lawrence - just a small correction. The address given above, gopher://micro.ifas.ufl.edu/11/NMR_Information_Server points to the NMR Information Server that has been established at Microbiology and Cell Science Department of the University of Florida - not NMRFAM. The server provides numerous links to scattered sources of information that are of interest to the NMR community. Among others, you will find a link to NMRFAM. Happy browsing, Marian === Marian Buszko (marian@micro.ifas.ufl.edu) Univ. of Florida, Microbiology From owner-structural-nmr@net.bio.net Wed Apr 20 23:00:00 1994 Path: biosci!OTTER.BIOCHEM.UBC.CA!mcintosh From: mcintosh@OTTER.BIOCHEM.UBC.CA (Lawrence McIntosh) Newsgroups: bionet.structural-nmr Subject: Re: What is this? Date: 20 Apr 1994 21:23:51 -0700 Organization: BIOSCI International Newsgroups for Molecular Biology Lines: 100 Sender: daemon@net.bio.net Distribution: bionet Message-ID: References: <2oi17u$b1n@news.u.washington.edu> NNTP-Posting-Host: net.bio.net With a little rooting around, there are lots of Biological and structural data bases, gophers, etc to be found. ************************************ A few useful gopher sites are: ncsa-xmosaic-hotlist-format-1 Default National Institutes of Health (NIH) Gopher gopher://gopher.nih.gov/11/molbio Sat Nov 27 10:06:51 1993 Molecular Biology Databases gopher://gopher.nih.gov/11/res Sat Nov 27 10:07:10 1993 Grants and Research Information gopher://merlot.welch.jhu.edu/11/biol-search Sat Nov 27 10:07:29 1993 Searching For Biologists (via Welchlab/Johns Hopkins) gopher://marvel.loc.gov/1 Sat Nov 27 10:09:22 1993 Library of Congress (LC MARVEL) gopher://merlot.welch.jhu.edu/11/Database-Searches Sat Nov 27 10:11:10 1993 Searches of many MolBio Databases (via Welchlab/Johns Hopkins) gopher://acsinfo.acs.org/1 Sat Nov 27 10:12:10 1993 American Chemical Society gopher://stis.nsf.gov/1 Sat Nov 27 10:15:23 1993 National Science Foundation (NSF) Gopher (STIS) gopher://vixen.cso.uiuc.edu/11/UI/DInfo/chemsch/nmr Sat Feb 12 16:39:25 1994 Nuclear Magnetic Resonance (NMR) gopher://gopher.nmrfam.wisc.edu/1 Thu Mar 17 18:20:13 1994 National Magnetic Resonance Facility at Madison (NMRFAM) gopher://micro.ifas.ufl.edu/11/NMR_Information_Server Fri Mar 18 12:10:47 1994 NMR Information Server **************************************** Also, you can get the Biologist's guide to Internet Resources from anonymous ftp to rtfm.mit.edu pub/usenet/news.answers/biology/guide ***************************************** A real gem is the National Center for Biotechnology Information (NCBI), offering on-line Entrez (including Medline, most data bases, etc), and the Blast and Retreive e-mail servers. Try info@ncbi.nlm.nih.gov or net-info@ncbi.nlm.nih.gov Also anonymous ftp to 130.14.25.1 for Entrez There is also a free News letter from ncbi. ****************************************** I hope that helps a bit, even if it only the tip of the electronic iceberg for information junkies. Lawrence McIntosh University of British Columbia On 14 Apr 1994, Angeline Kantola wrote: > In article <1994Apr12.093021.8263@uts.uni-c.dk>, > Mogens Kjaer wrote: > >This newsgroup popped up im my newsreader a few days ago, but noone have > >posted anything so far - is there something wrong, or are there nobody > >out there working with structural NMR? > > Sure there are! > > OK, I've got something to ask. I'm a graduate student in Biochemistry, > studying protein structure via (surprise) NMR. Recently I found out about > and poked around in the NMRFAM gopher site out of the University of > Wisconsin; a variety of NMR-related programs are accessible through this > site. (Maybe someone who's affiliated with the NMR center at the other UW > could fill us in a little more about NRMFAM?) Aside from the PDB site > itself, are there other gopher or ftp sites related to protein structure > or strucutural NMR? > > --Angie From owner-structural-nmr@net.bio.net Wed Apr 20 23:00:00 1994 Path: biosci!OTTER.BIOCHEM.UBC.CA!mcintosh From: mcintosh@OTTER.BIOCHEM.UBC.CA (Lawrence McIntosh) Newsgroups: bionet.structural-nmr Subject: info Date: 20 Apr 1994 21:53:25 -0700 Organization: BIOSCI International Newsgroups for Molecular Biology Lines: 101 Sender: daemon@net.bio.net Distribution: bionet Message-ID: NNTP-Posting-Host: net.bio.net With a little rooting around, there are lots of Biological and structural data bases, gophers, etc to be found. ************************************ A few useful gopher sites are: ncsa-xmosaic-hotlist-format-1 Default National Institutes of Health (NIH) Gopher gopher://gopher.nih.gov/11/molbio Sat Nov 27 10:06:51 1993 Molecular Biology Databases gopher://gopher.nih.gov/11/res Sat Nov 27 10:07:10 1993 Grants and Research Information gopher://merlot.welch.jhu.edu/11/biol-search Sat Nov 27 10:07:29 1993 Searching For Biologists (via Welchlab/Johns Hopkins) gopher://marvel.loc.gov/1 Sat Nov 27 10:09:22 1993 Library of Congress (LC MARVEL) gopher://merlot.welch.jhu.edu/11/Database-Searches Sat Nov 27 10:11:10 1993 Searches of many MolBio Databases (via Welchlab/Johns Hopkins) gopher://acsinfo.acs.org/1 Sat Nov 27 10:12:10 1993 American Chemical Society gopher://stis.nsf.gov/1 Sat Nov 27 10:15:23 1993 National Science Foundation (NSF) Gopher (STIS) gopher://vixen.cso.uiuc.edu/11/UI/DInfo/chemsch/nmr Sat Feb 12 16:39:25 1994 Nuclear Magnetic Resonance (NMR) gopher://gopher.nmrfam.wisc.edu/1 Thu Mar 17 18:20:13 1994 National Magnetic Resonance Facility at Madison (NMRFAM) gopher://micro.ifas.ufl.edu/11/NMR_Information_Server Fri Mar 18 12:10:47 1994 NMR Information Server **************************************** Also, you can get the Biologist's guide to Internet Resources from anonymous ftp to rtfm.mit.edu pub/usenet/news.answers/biology/guide ***************************************** A real gem is the National Center for Biotechnology Information (NCBI), offering on-line Entrez (including Medline, most data bases, etc), and the Blast and Retreive e-mail servers. Try info@ncbi.nlm.nih.gov or net-info@ncbi.nlm.nih.gov Also anonymous ftp to 130.14.25.1 for Entrez There is also a free News letter from ncbi. ****************************************** I hope that helps a bit, even if it only the tip of the electronic iceberg for information junkies. Lawrence McIntosh University of British Columbia On 14 Apr 1994, Angeline Kantola wrote: > In article <1994Apr12.093021.8263@uts.uni-c.dk>, > Mogens Kjaer wrote: > >This newsgroup popped up im my newsreader a few days ago, but noone have > >posted anything so far - is there something wrong, or are there nobody > >out there working with structural NMR? > > Sure there are! > > OK, I've got something to ask. I'm a graduate student in Biochemistry, > studying protein structure via (surprise) NMR. Recently I found out about > and poked around in the NMRFAM gopher site out of the University of > Wisconsin; a variety of NMR-related programs are accessible through this > site. (Maybe someone who's affiliated with the NMR center at the other UW > could fill us in a little more about NRMFAM?) Aside from the PDB site > itself, are there other gopher or ftp sites related to protein structure > or strucutural NMR? > > --Angie From owner-structural-nmr@net.bio.net Fri Apr 22 23:00:00 1994 Path: biosci!rutgers!usc!howland.reston.ans.net!pipex!lyra.csx.cam.ac.uk!mole.bio.cam.ac.uk!smb18 From: smb18@mole.bio.cam.ac.uk (Simon Brocklehurst (Bioc)) Newsgroups: bionet.structural-nmr Subject: Protein stereochemistry Message-ID: <2p5mc5$p83@lyra.csx.cam.ac.uk> Date: 21 Apr 94 11:00:21 GMT Organization: University of Cambridge, England Lines: 27 NNTP-Posting-Host: mole.bio.cam.ac.uk Following on from some of the discussion about precision and accuracy of DNA structures, maybe this would be an appropriate time to attempt to start a discussion about NMR protein structures. One problem with the force fields for solving protein structures is that we always get covalent geometry that is too tight. Until recently, this was probably unavoidable because it was difficult to get enough restraints to define the structure well. But by using doubled labelled protein we can now get a LOT of restraints. Does anyone think it's important that we properly address the problem of allowing appropriate distortions of covalent geometry? If so, what do people think is a high enough number of restraints / low enough ensemble r.m.s.d. to start making this doable/useful/possible...? -- Simon !----------------------------------------------------------------------- Simon M. Brocklehurst Cambridge Centre for Molecular Recognition Department of Biochemistry University of Cambridge Cambridge UK E-mail: s.m.brocklehurst@bioc.cam.ac.uk From owner-structural-nmr@net.bio.net Sat Apr 23 23:00:00 1994 Path: biosci!rutgers!uwvax!uchinews!vixen.cso.uiuc.edu!cs.uiuc.edu!wupost!monsanto.com!alex.monsanto.com!wchutt From: wchutt@alex.monsanto.com (Bill C Hutton) Newsgroups: bionet.structural-nmr Subject: Re: Welcome II Message-ID: <1994Apr20.153610.20307@tin.monsanto.com> Date: 20 Apr 94 15:36:10 GMT Sender: news@tin.monsanto.com (USENET News System) Organization: Monsanto Company Lines: 31 X-Newsreader: TIN [version 1.2 PL0] David wirtes: : But I am in agreement that *some* interesting information can be gathered. I : just don't think the 'state of the art' is sufficiently advanced to give : the old '0.2 A RMS' structures that protein NMR types have come to demand. =^) Which begs the question: what is going on in these systems that is different. Of course there are fewer constraints. Current thinking is something on the order of "lots of imprecise constraints are ok to get a good structure...just get as many as you can". Do the simplifications of relaxation theory break down in DNA/RNA but not in proteins? Are the simplifications similiarly poor in proteins but the large number of constraints save the day? Will anyone ever be able to simulate the time-dependent Hamiltonians for real spin systems such that when you subtract an empirical NOESY spectrum from a simulated spectrum the residual is zero (this is trivial for the time independent terms)? Can anyone rule out the current state of H-H relaxation theory is not up to the job? What about CSA-dipole cross corelation, dipole-dipole cross correlation, the efects of J coupling and what about second order coupling? These relaxation effects could be used to define a new type of constraint, a geometric or spatial constraint (what angle or angles does the spin system - note I purposely did not use the term spin pair- have with respect to the major axis of molecular tumbling?). Matrix approaches will never work well if the underlying assumptions about the form of the spectral density functions is incomplete. This NOT to say the current methodology of no use... it obviously is. The question is will it ever get better (I'm still looking for that zero residual spectrum) without more sophisticated relaxation spectral densities. What will it take to get us to the next level of precision? If I only had the time and $...... From owner-structural-nmr@net.bio.net Sat Apr 23 23:00:00 1994 Path: biosci!rutgers!uwvax!uchinews!vixen.cso.uiuc.edu!cs.uiuc.edu!wupost!cs.utexas.edu!howland.reston.ans.net!pipex!warwick!lyra.csx.cam.ac.uk!mole.bio.cam.ac.uk!arcr1 From: arcr1@mole.bio.cam.ac.uk (Andy Raine (Bioc)) Newsgroups: bionet.structural-nmr Subject: Re: DNA duplex structures by NMR Message-ID: <2p5fes$mdo@lyra.csx.cam.ac.uk> Date: 21 Apr 94 09:02:20 GMT References: <9404191522.AA03119@primer.med.yale.edu> <2p2ob1$n2t@lyra.csx.cam.ac.uk> Distribution: bionet Organization: U. of Cambridge, England Lines: 29 NNTP-Posting-Host: mole.bio.cam.ac.uk Oops. Here is a follow-up to my own follow-up: In article <2p2ob1$n2t@lyra.csx.cam.ac.uk>, I wrote: |> I may be missing something here, but this worries me a little. Do structures |> calculated using the "James weighting" agree with your experimental data better or |> worse than those calculated with a more conventional protocol? If the fit is no |> better, then an ensemble with a greater RMSD would be a more honest interpretation |> of your data. |> |> If the structures fit the data better, as well as giving a lower RMSD then there |> is no problem of course! I should have said: If the structures fit the data better, and the stereochemical parameters are as good, as well as giving a lower RMSD then there is no problem of course! Andrew -------------------------------------------------------------------------------- Dr. Andrew Raine +44 223 333744 Department of Biochemistry or 333499 University of Cambridge Tennis Court Road Cambridge CB2 1QW United Kingdom -------------------------------------------------------------------------------- From owner-structural-nmr@net.bio.net Tue Apr 26 23:00:00 1994 Path: biosci!daresbury!not-for-mail From: driscoll@bioch.ox.ac.uk (Paul Driscoll) Newsgroups: bionet.structural-nmr Subject: NMR POSITIONS - LONDON Date: 27 Apr 1994 19:29:53 +0100 Lines: 71 Sender: daemon@mserv1.dl.ac.uk Distribution: bionet Message-ID: <2pmav1$1kl@mserv1.dl.ac.uk> content-length: 2545 Original-To: str-nmr@dl.ac.uk Please post/distribute... NMR SPECTROSCOPY OF PROTEINS POSTDOCTORAL FELLOWSHIP POSITIONS LUDWIG INSTITUTE OF CANCER RESEARCH and UNIVERSITY COLLEGE LONDON - DEPARTMENT OF BIOCHEMISTRY AND MOLECULAR BIOLOGY We are looking for TWO people to join a new biological NMR spectroscopy laboratory that is in the process of construction at University College London (completion Summer 1994). Successful applicants will be involved in the study of protein structure, dynamics and ligand binding of signal transduction molecules (e.g. PI3 Kinase) that are the focus of the research of Prof. Mike D. Waterfield FRS (Director of the Ludwig Institute). The laboratory is to include two brand new high field Varian UnityPLUS spectrometers (one 500 MHZ and one 600 MHz) including triple resonance and pulse field gradient capabilities. The laboratory will be equipped with a suite of Silicon Graphics Indy and Indigo workstations and will share a four processor SGI Challenge server with the group Prof. Janet Thornton. A new 'wet biochemistry' laboratory will also be established and it will be possible to tap into the extensive biochemistry resources provided by the Ludwig Institute. Applicants should either: Have previous experience of development of NMR methods, or the application of NMR techniques to the study of protein structure and function; or Have previous experience of molecular biology and/or large scale protein expression/purification skills and be able to demonstrate an appreciation of the requirements of biological NMR studies. Salaries (subject to negotiation) will be related to the MRC non-clinical Scientific Salary Scale (approximately £15795 -> £25107) plus the LICR London weighting (£2750 per annum). The start date is negotiatable. Applicants should forward a copy of their CV and the names of three referees to Mr. Richard Anning, Ludwig Institute of Cancer Research, Courtauld Building, 91 Riding House Street, London W1P 8BT, U.K. Informal enquiries may be made to Dr. Paul Driscoll (see details below). [....This is a repeat of the advertisement published in 7th April 1994 issue of the journal Nature....] ----------------------------------------------------------------- Paul C. Driscoll Royal Society University Research Fellow (Currently at...) Department of Biochemistry, University of Oxford, South Parks Road, Oxford OX1 3QU. UK. tel. (44) 865 275799 fax. (44) 865 275259 E-mail: driscoll@bioch.ox.ac.uk From owner-structural-nmr@net.bio.net Tue Apr 26 23:00:00 1994 Path: biosci!daresbury!trane.uninett.no!nac.no!nntp-oslo.uninett.no!usenet From: ibfls@nlh10.nlh.no Newsgroups: bionet.structural-nmr Subject: Macintosh software Date: 27 Apr 1994 12:35:06 GMT Organization: agricultural university of norway Lines: 4 Distribution: world Message-ID: <2plm5q$jj2@ratatosk.uninett.no> NNTP-Posting-Host: ibf-ls.nlh.no Does anyone have any experience with any Macintosh software within protein NMR? Lars Skjeldal e-mail adrss: ibfls@nlh10.nlh.no From owner-structural-nmr@net.bio.net Tue Apr 26 23:00:00 1994 Newsgroups: bionet.structural-nmr Path: biosci!daresbury!trane.uninett.no!sunic!EU.net!howland.reston.ans.net!torn!nott!cunews!superior!mreynold From: mreynold@superior.carleton.ca (mark reynolds) Subject: PC Spin Sim Message-ID: Sender: news@cunews.carleton.ca (News Administrator) Organization: Carleton University X-Newsreader: TIN [version 1.2 PL0] Date: Wed, 27 Apr 1994 14:37:31 GMT Lines: 5 Does anybody know of a program for the PC which simulates spin patterns? For example put in shift values and coupling constants, and the program gives a theoretical spectrum. Marcus From owner-structural-nmr@net.bio.net Tue Apr 26 23:00:00 1994 Path: biosci!BRESSE.CBS.UNIV-MONTP1.FR!jrene From: jrene@BRESSE.CBS.UNIV-MONTP1.FR (Jean Rene Alattia) Newsgroups: bionet.structural-nmr Subject: spectral density function Date: 27 Apr 1994 07:25:57 -0700 Organization: BIOSCI International Newsgroups for Molecular Biology Lines: 18 Sender: daemon@net.bio.net Distribution: bionet Message-ID: <9404271415.AA27019@bresse.cbs.univ-montp1.fr> NNTP-Posting-Host: net.bio.net . . . >Matrix approaches will never work well if the underlying assumptions about the >form of the spectral density functions is incomplete. This NOT to say the . . . Bill, you said it all! Molecules in solution do move (internally and globaly) >If I only had the time and $...... That's a real problem, isn't it ?! From owner-structural-nmr@net.bio.net Tue Apr 26 23:00:00 1994 Newsgroups: bionet.structural-nmr Path: biosci!daresbury!trane.uninett.no!sunic!EU.net!howland.reston.ans.net!torn!watserv2.uwaterloo.ca!peptidarus.UWaterloo.ca!ltaylor From: ltaylor@peptidarus.UWaterloo.ca (Lorne Taylor) Subject: NOE Volumes (peptides) Message-ID: Sender: news@watserv2.uwaterloo.ca Organization: University of Waterloo Date: Wed, 27 Apr 1994 13:58:19 GMT Lines: 30 OK...this newsgroup has been open long enough for the naive and inexperienced to start to bother everyone, so here I go... I've assigned all the protons in a 16 and a 24 residue peptide in DMSO. Now I want to try and get distance constraints from the NOE volumes. So: 1) If I want to try to extract buildup curves for the NOE's, how do I pick a baseline (below is noise/above is data) that is the same for all mixing times. Is the exact baseline value crucial? I suspect it is as even small variations in the level could cause big volume changes. 2) The resolution is fairly high (2k real X 512 expts) and I've zero filled to 2K X 2K. My problem is some peaks are quite well resolved and some are not (ie SOME amide to alpha protons show a DQCOSY-like square array (AX) and others are a single relatively broad peak). Since the well resolved peaks have much less area than unresolved ones would not the volumes also be uncomparable?? 3) I've some qualitative idea of the conformation (ie NH-NH (i,i+1)) NOE in the presence of stronger (NH-alpha H (i,i+1)) which I interpret as a peptide that spends some time as a helix but more time in unordered conformations. If I put mutually exclusive contraints such as these into a distance geometry or restrained MD program, would I get a meaningful answer??? Thanks to all in advance; Lorne From owner-structural-nmr@net.bio.net Tue Apr 26 23:00:00 1994 Path: biosci!TOME.CBS.UNIV-MONTP1.FR!jrene From: jrene@TOME.CBS.UNIV-MONTP1.FR (Jean Rene Alattia) Newsgroups: bionet.structural-nmr Subject: non contacts Date: 27 Apr 1994 14:39:24 -0700 Organization: BIOSCI International Newsgroups for Molecular Biology Lines: 20 Sender: daemon@net.bio.net Distribution: bionet Message-ID: <199404272139.OAA11115@net.bio.net> NNTP-Posting-Host: net.bio.net >Does anybody care to continue? Yes but to raise another question: what about the use of "non contacts" to constrain molecules (i.e setting the lower bound of the distance between two nuclei to , say 5 A, and the upper bound to an arbitrarily large limit when no NOEs are observed between those nuclei)? Has anyone any examples about the use of such constraints or justifications for such use? I think one has to be sure that the structure is sufficiently rigid (can we really say a structure is rigid ?!!) or let us say well defined before including "non contacts" in structure calculations. ######################################################################### # Jean-Rene ALATTIA # # # Centre de Biochimie Structurale # TEL: (33) 67 04 34 33 # # Faculte de Pharmacie # FAX: (33) 67 52 96 23 # # 15, Avenue Charles Flahault # # # 34060 Montpellier Cedex 1 # jrene@cbs.univ-montp1.fr # ######################################################################### From owner-structural-nmr@net.bio.net Tue Apr 26 23:00:00 1994 Path: biosci!PICASSO.UCSF.EDU!ulyanov From: ulyanov@PICASSO.UCSF.EDU (Nikolai Ulyanov) Newsgroups: bionet.structural-nmr Subject: Re: Welcome II Date: 27 Apr 1994 12:56:55 -0700 Organization: BIOSCI International Newsgroups for Molecular Biology Lines: 84 Sender: daemon@net.bio.net Distribution: bionet Message-ID: <199404271953.MAA17226@picasso.ucsf.EDU> NNTP-Posting-Host: net.bio.net wchutt@alex.monsanto.com (Bill C Hutton) writes: > David wirtes: > > : But I am in agreement that *some* interesting information can be gathered. I > : just don't think the 'state of the art' is sufficiently advanced to give > : the old '0.2 A RMS' structures that protein NMR types have come to demand. =^) > > Which begs the question: what is going on in these systems that is different. > Of course there are fewer constraints. Current thinking is something on the > order of "lots of imprecise constraints are ok to get a good structure...just > get as many as you can". Do the simplifications of relaxation theory break down > in DNA/RNA but not in proteins? Are the simplifications similiarly poor in > proteins but the large number of constraints save the day? Will anyone ever be > able to simulate the time-dependent Hamiltonians for real spin systems such that > when you subtract an empirical NOESY spectrum from a simulated spectrum the > residual is zero (this is trivial for the time independent terms)? Can anyone > rule out the current state of H-H relaxation theory is not up to the job? What > about CSA-dipole cross corelation, dipole-dipole cross correlation, the efects > of J coupling and what about second order coupling? These relaxation effects > could be used to define a new type of constraint, a geometric or spatial > constraint (what angle or angles does the spin system - note I purposely did > not use the term spin pair- have with respect to the major axis of molecular > tumbling?). > > Matrix approaches will never work well if the underlying assumptions about the > form of the spectral density functions is incomplete. This NOT to say the > current methodology of no use... it obviously is. The question is will it ever > get better (I'm still looking for that zero residual spectrum) without > more sophisticated relaxation spectral densities. What will it take to get us > to the next level of precision? > > If I only had the time and $...... > Bill brought up a lot of interesting and not easy questions, and, frankly, I wanted to hear somebody else answering them. But nobody is doing that so far, so I decided to start with the easier ones, mainly to stir up the discussion.. > ... Do the simplifications of relaxation theory break down > in DNA/RNA but not in proteins? Are the simplifications similiarly poor in > proteins but the large number of constraints save the day? The simplifications are, of course, the same.. I think, not the number of restraints save the day with proteins, but their information content. If, say, you have a cross peak btw residues 1 and 100 in a polypeptide, it gives you a lot of important information, even if you don't have anything else. And it does not really matter if the corresponding distance is 2 A or 5 A; you know that there is a tertiary interaction in the protein (at least part of the time). In the case of standard DNA duplexes, the answer is known at such a level of detalization for decades: it is the B-form; with a very few exceptions you can predict all cross peaks without doing any experiments. What is interesting here, are very subtle (compared to proteins) structural details: small bends, small under- or over-twistings, etc. Naturally, one needs much more accurate methods to get such an information. And of course, possible internal motions can be a major obstacle here: > Matrix approaches will never work well if the underlying assumptions about the > form of the spectral density functions is incomplete. And this could create severe problems for peptides as well; see, e.g., ltaylor@peptidarus.UWaterloo.ca (Lorne Taylor) writing: > I've assigned all the protons in a 16 and a 24 residue peptide in DMSO. > <...> > ... If I put mutually exclusive contraints > such as these into a distance geometry or restrained MD program, > would I get a meaningful answer??? > Not likely, I think.. Nevertheless, there is still something that might be done if, e.g., internal motion is slow compared to overall tau-c. One way to go would be to generate MD-TAR ensembles where restraints must be satisfied on the time-average basis (Schmitz et al. 1993 JMB 234:373 for DNA, and Torda et al. 1990 JMB 214:223 for tendamistat). Another way is to try to identify mutually exclusive subsets of restraints and refine separate conformers (Blackledge et al. 1993 Biochemistry 32:10960). Does anybody care to continue? Nick From owner-structural-nmr@net.bio.net Wed Apr 27 23:00:00 1994 Path: biosci!RISVAX.ROWLAND.ORG!HOCH From: HOCH@RISVAX.ROWLAND.ORG (Jeffrey C. Hoch) Newsgroups: bionet.structural-nmr Subject: Reply to Paul Driscoll on restraint violations Date: 28 Apr 1994 13:15:49 -0700 Organization: BIOSCI International Newsgroups for Molecular Biology Lines: 83 Sender: daemon@net.bio.net Distribution: bionet Message-ID: <940428161613.2925@RISVAX.ROWLAND.ORG> NNTP-Posting-Host: net.bio.net Paul: The problem with violated experimental restraints is less a problem with simulated annealing than the force fields. We are trying to solve a constrained optimization problem of the following type: minimize E subject to the constraint that the structure is consistent with the experimental data. The constraint is usually quantified by the requirement that a constraint statistic C be less than some value C(0) (e.g. the rms violation). The standard method for solving this type of problem is to turn it into an unconstrained optimization by the introduction of a new objective function O = E + lambda * C where E is the empirical energy function, C is the constraint statistic, and lambda is a "Lagrange multiplier", effectively the weight applied to the constraint statistic. We then minimize O. There will be solutions whenever the gradient of O vanishes. I distinguish two types of solutions: "trivial" solutions, which occur when grad E = grad C = zero, and "non-trivial" solutions, which occur when grad E = - lambda * grad C. The existence of a trivial solution indicates that there is a local minimum of the potential energy function that statisfies the experimental constraints; in principle we don't need the experimental restraints if we can identify that local minimum of the force field. Of course there are a number of problems; even if the force field is sufficiently accurate for such a local minimum to exist, there is no guarantee that there is a unique solution. If we are fortunate enough that the correct solution is unique and corresponds to a global minimum of the potential function, we might not have an easy time finding it in a sea of local minima. If the force field is not sufficiently accurate, we will only find non-trivial solutions. This seems almost undoubtably to be the case, and manifests itself in a number of ways. For example, MD simulations of proteins that start from a crystal structure almost always drift from the starting point. The reasons are myriad, as well, including errors in the parameterization of the force field, neglect or inaccurate treatment of solvent, differences in time scales (which time-averaged restraints seek to address), and so on. The form of the restraints (flat or harmonic) means that in order to get a non-zero gradient, there must be a violation. There does appear to be one interesting case of an MD simulation of a biomolecule that did not stray from a configuration that satisfied the experimental constraints when they were "turned off". That is for a DNA oligomer, reported by Withka, Swaminathan, and Beveridge in Science a couple of years ago. I don't know of any similar cases for a protein. If we must be content with non-trivial solutions to our optimization problem, how do we proceed? Well, we choose lambda (the weight applied to the experimetal constraints) so that the value of the constraint statistic for a solution is comparable to our estimate of the uncertainty inherent in the experimental constraints. We don't want to fit the data too well so that we are not guilty of overfitting. We do have >some< faith in our potential function, so we usually insist that in addition to fitting the experimental data the potential energy should be "reasonable". There is room for cleverness here, since not all the terms in the potential energy function are equally accurate. Again, if we are too insistent on obtaining a low potential energy, we may be guilty of over-fitting if we exceed the accuracy of the potential function. There is an analogy between the non-trivial solutions to our optimization problem and a phenomenum that solid-state physicists call, appropriately enough, "frustration". This occurs when it is not possible to simultaneously minimize all of the components of the energy. However we choose to find these solutions, whether by simulated annealing, homotopy methods, genetic algorithms, or whatever, we require that they be feasible, in that they account for the experimental data, and are at the same time plausible, in that they have reasonable geometries (or low potential energies). To be too insistent on either of these criteria is to risk over-fitting; one characteristic of over-fitting is that the distribution of structures will be artificially narrow. Which brings us back to your point about "RMSD chasing". I agree completely that not enough attention has been given to the approprate *lower* bound on the RMSD. Enough for now, Jeffrey Hoch Rowland Institute for Science Cambridge, Massachusetts hoch@rowland.org From owner-structural-nmr@net.bio.net Wed Apr 27 23:00:00 1994 Path: biosci!daresbury!not-for-mail From: driscoll@bioch.ox.ac.uk (Paul Driscoll) Newsgroups: bionet.structural-nmr Subject: Re: Welcome II Date: 28 Apr 1994 11:25:12 +0100 Lines: 47 Sender: daemon@mserv1.dl.ac.uk Distribution: bionet Message-ID: <2po2u8$sf8@mserv1.dl.ac.uk> content-length: 2365 Original-To: str-nmr@dl.ac.uk Surely the point is that in proteins the 'quality' - accuracy and precison - of the structure obtained from NMR restraints is spatially dependent. In other words, the core of the protein is much more likely to be of higher 'quality' than the surface (where the density and cooperativity of experimental restraints is almost certainly lower). I expect there are parallels in DNA solution structure determination. Perhaps progess in the future will come from superior characterisation of dihedral angles (from improved measurements of spin-spin coupling constants) - although even these will suffer from model-dependent interpretation in instances of time-averaged conformational fluctuations. Almost every protein I have ever studied by NMR has shown evidence of slow conformational equilibria and therefore the solution 'structure' must be disordered. The beauty of NMR (over say X-ray diffraction) is that it provides a much more transparent avenue to study the dynamics processes in a protein. RMSD-chasers are missing the point, it seems to me. NMR exquisitely provides information about molecular motions within certain timescales and it is the inclusion of experimentally determined dynamic parameters (e.g S**2 order parameters) into the characterisation of NMR structures that lies at the frontier of our technology. Whether we can actually map spectral densities seems to be an open question. I have another question to put to those doing dynamical simulated annealing. If you have a completely accurate set of distances for a protein and you apply them in a SA solution structure protocol, you will never get a zero NOE restraint residual (at least you don't in X-plor) because of the partitioning of energy amongst the different terms of the force-field. So even though your distances are all in theory completely and internally consistent you will never actually recover the 'real' structure, since there will always be an NOE residual (however small). What does this say about the SA method? Comments appreciated. P.s. Is there anyone out ther at all getting the newsgroup via USENET in the U.K. 'cos we shurely are not? Paul C. Driscoll Royal Society University Research Fellow Department of Biochemistry, University of Oxford, South Parks Road, Oxford OX1 3QU. UK. tel. (44) 865 275799 fax. (44) 865 275259 E-mail: driscoll@bioch.ox.ac.uk From owner-structural-nmr@net.bio.net Wed Apr 27 23:00:00 1994 Path: biosci!BIONET.IG.COM!kristoff From: kristoff@BIONET.IG.COM (David Kristofferson) Newsgroups: bionet.structural-nmr Subject: test of str-nmr@net.bio.net Date: 28 Apr 1994 15:59:40 -0700 Organization: BIOSCI International Newsgroups for Molecular Biology Lines: 2 Sender: daemon@net.bio.net Distribution: bionet Message-ID: NNTP-Posting-Host: net.bio.net test of str-nmr@net.bio.net From owner-structural-nmr@net.bio.net Wed Apr 27 23:00:00 1994 Path: biosci!biosci!not-for-mail From: kristoff@net.bio.net (David Kristofferson) Newsgroups: bionet.structural-nmr Subject: Re: Hello? Date: 28 Apr 1994 15:58:53 -0700 Organization: BIOSCI International Newsgroups for Biology Lines: 1 Distribution: bionet Message-ID: <2ppf3d$koh@net.bio.net> References: <2pioib$dat@news.uni-c.dk> NNTP-Posting-Host: net.bio.net This is a test of bionet.structural-nmr. From owner-structural-nmr@net.bio.net Wed Apr 27 23:00:00 1994 Path: biosci!daresbury!trane.uninett.no!sunic!news.uni-c.dk!unidhp.uni-c.dk!carlmk From: carlmk@unidhp.uni-c.dk (Mogens Kjaer) Newsgroups: bionet.structural-nmr Subject: Hello? Date: 26 Apr 1994 09:57:31 GMT Organization: News Server at UNI-C, Danish Computing Centre for Research and Education. Lines: 19 Message-ID: <2pioib$dat@news.uni-c.dk> NNTP-Posting-Host: unidhp.uni-c.dk Summary: Is this working? X-Newsreader: TIN [version 1.2 PL1] I've waited and waited and waited for someone to post anything in this newsgroup. So far, I've only seen a response from another Danish guy. Is this an empty newsgroup, or is there something wrong with our news server? Mogens -- Mogens Kjaer Pronto Software - Development and Distribution Carlsberg Research Center Gamle Carlsberg Vej 10 DK-2500 Valby Denmark Phone: + 45 33 27 53 25 Fax: + 45 33 27 47 08 Email: carlmk@unidhp.uni-c.dk From owner-structural-nmr@net.bio.net Wed Apr 27 23:00:00 1994 Path: biosci!daresbury!not-for-mail From: David Kristofferson Newsgroups: bionet.structural-nmr Subject: test of str-nmr@daresbury.ac.uk Date: 29 Apr 1994 00:01:56 +0100 Lines: 2 Sender: daemon@mserv1.dl.ac.uk Distribution: bionet Message-ID: <2ppf94$dek@mserv1.dl.ac.uk> Original-To: str-nmr@dl.ac.uk test of str-nmr@daresbury.ac.uk From owner-structural-nmr@net.bio.net Wed Apr 27 23:00:00 1994 Newsgroups: bionet.structural-nmr Path: biosci!daresbury!trane.uninett.no!sunic!uts!news From: cc@charon.dfh.dk Subject: Re: PC Spin Sim Reply-To: cc@charon.dfh.dk Organization: Royal Danish School of Pharmacy Date: Thu, 28 Apr 1994 09:39:37 GMT Message-ID: <1994Apr28.093937.24344@uts.uni-c.dk> Lines: 31 X-Newsreader: IBM NewsReader/2 v1.00 References: Sender: news@uts.uni-c.dk (News) Nntp-Posting-Host: cornetto.dfh.dk Lines: 32 In , mreynold@superior.carleton.ca (mark reynolds) writes: >Does anybody know of a program for the PC which simulates spin >patterns? For example put in shift values and coupling constants, and >the program gives a theoretical spectrum. > >Marcus Professor Schaumburg at the University of Copenhagen has ported MIMER (QCPE No. 394, QCPE 13 (1981)) to the IBM PC. It will run on anything from an 8088 equipped with an '87 (mandtory) and upwards. It consists of a simulation part, MIMER nad a drawing part, ARTIST. PC-version accpets up to 7 nuclei, 1 of these X-approximated as far as I remember. Further references: O. Manscher, K. Schaumburg and J.P. Jacobsen. Acta Chem. Scand. ser A. _35_, (1981), 13-24. If interested You should contact him via snail-mail at: Professor K. Schaumburg, Department of Chemical Physics Chemical Laboratory 5 Universitetsparken 2 DK 2100 CPH P.S. Its a FORTRAN program with FORTRAN I/O... but it works. We're contemplating porting it to OS/2. Claus Cornett Institute for Analytical and Pharmaceutical Chemistry The Royal Danish School of Pharmacy Universitetsparken 2 DK 2100 Copenhagen Fax. +45 35375744 From owner-structural-nmr@net.bio.net Wed Apr 27 23:00:00 1994 Path: biosci!PICASSO.UCSF.EDU!schmitz From: schmitz@PICASSO.UCSF.EDU Newsgroups: bionet.structural-nmr Subject: RE: non contacts Date: 27 Apr 1994 17:48:54 -0700 Organization: BIOSCI International Newsgroups for Molecular Biology Lines: 32 Sender: daemon@net.bio.net Distribution: bionet Message-ID: <199404280049.RAA23205@dali.ucsf.edu> NNTP-Posting-Host: net.bio.net Jean-Rene Alattia writes: > Has anyone any examples about the use of such constraints or justifications > for such use? I think one has to be sure that the structure is sufficiently > rigid (can we really say a structure is rigid ?!!) or let us say well defined > before including "non contacts" in structure calculations I have had some good experiences with "non contact" restraints. In the case of a DNA octamer two THY were almost completely overlapped. Very few restraints were available and my rMD simulations did not converge when I started fro A-DNA. Instead, the molecule got trapped in "weird" structures, particularly for those THYs. Here, it was straightforward to calculate the theoretical NOE spectrum for those weird structures which would have a number of at least medium NOEs that I unequivocally had not observed. Hence, I made up non contact restraints by assigning lower bounds between 4 and 5 A. Upper bounds dont matter, I assigned zero force-constants anyway. In this case, it was clear that nothing weird was going on with those protons since they had normal linewidth, they were just overlapped. However, there might be other reasons why peaks are not observed, esp. in systemes with differential mobilities (loops, proteins in general). IMHO, these indirect restraints can be pretty powerful but caution is advised. We are thinking about using them in a last refinement step, when they overall structure has been determined and predicted NOEs are not in the spectra. Uli ><><><><><><><><><><><><><><><><><><> Uli Schmitz UC,San Francisco Dept. Pharmaceutical Chemistry Phone (415) 476 4378; Fax (415) 476 0688; schmitz@picasso.ucsf.edu ><><><><><><><><><><><><><><><><><><> From owner-structural-nmr@net.bio.net Wed Apr 27 23:00:00 1994 Newsgroups: bionet.structural-nmr Path: biosci!daresbury!trane.uninett.no!sunic!EU.net!sun4nl!news.nic.surfnet.nl!ruu.nl!RUUCI12!leef From: leef@nmr.chem.ruu.nl (Bas Leeflang) Subject: Re: non contacts Message-ID: <1994Apr28.080508.8789@cc.ruu.nl> Sender: leef@RUUCI12 (Bas Leeflang) Organization: ACCU References: <199404272139.OAA11115@net.bio.net> Distribution: bionet Date: Thu, 28 Apr 1994 08:05:08 GMT Lines: 26 In article <199404272139.OAA11115@net.bio.net>, jrene@TOME.CBS.UNIV-MONTP1.FR (Jean Rene Alattia) writes: ... what about the use of "non contacts" |> to constrain molecules (i.e setting the lower bound of the distance between |> two nuclei to , say 5 A, and the upper bound to an arbitrarily large limit |> when no NOEs are observed between those nuclei)? |> Has anyone any examples about the use of such constraints or justifications |> for such use? I think one has to be sure that the structure is sufficiently |> rigid (can we really say a structure is rigid ?!!) or let us say well defined |> before including "non contacts" in structure calculations. I would say that 'non-contacts' could be included in time-averaged aproaches. Here a structure does not need to be rigid, but on average a distance must be larger than the lowerbound. I wonder, however, if non-contacts will contribute much to the structure. Are there many/any non-contact violations in structures presently under study???? That is short distances in the model and the absence of a NOE peak withou an obvious reason (presaturation). ---------------------------------------------------------------------- Bas R. Leeflang _/ _/ _/ _/ _/_/_/ Department of the Bijvoet Center _/_/ _/ _/_/ _/_/ _/ _/ for Biomolecular Research _/ _/ _/ _/ _/ _/ _/_/_/ Padualaan 8, 3584 CH Utrecht _/ _/_/ _/ _/ _/ _/ Tel. : int+31.30.533295 _/ _/ _/ _/ _/ _/ Fax : int+31.30.537623 Email: leef@nmr.chem.ruu.nl From owner-structural-nmr@net.bio.net Wed Apr 27 23:00:00 1994 Newsgroups: bionet.structural-nmr Path: biosci!daresbury!trane.uninett.no!sunic!news.funet.fi!ousrvr.oulu.fi!eha From: eha@phoenix.oulu.fi (Esa Haapaniemi) Subject: Re: PC Spin Sim Message-ID: <1994Apr28.064545.4283@ousrvr.oulu.fi> Sender: news@ousrvr.oulu.fi Organization: University of Oulu, Finland X-Newsreader: TIN [version 1.2 PL1] References: Date: Thu, 28 Apr 1994 06:45:45 GMT Lines: 22 : Does anybody know of a program for the PC which simulates spin : patterns? For example put in shift values and coupling constants, and : the program gives a theoretical spectrum. There are several such programs available, but THE BEST IMO is Perch. It is commercial package from Perch project University of Kuopio, Finland, but there is fully usable demoversion of it available. Only restrictions on the demo are that it can handle only up to 8 spins, when the full can handle 10 or more. Also plenty of importer routines for measured spectra are grippled, but they are valuable only if you want to make iterative analysis from measured spectrum. The demoversion in available from ftp.funet.fi under some directory that I don't remember the name of ;I That demoversion can be used either from windows or DOS, and it supports printing, laocon type iterative analysis, MLD, ... Esa Haapaniemi University of Oulu Department of Chemistry Finland From owner-structural-nmr@net.bio.net Wed Apr 27 23:00:00 1994 Newsgroups: bionet.structural-nmr Path: biosci!daresbury!trane.uninett.no!sunic!EU.net!sun4nl!news.nic.surfnet.nl!ruu.nl!RUUCI12!leef From: leef@RUUCI12.NoSubdomain.NoDomain (Bas Leeflang) Subject: Re: NOE Volumes (peptides) Message-ID: <1994Apr28.075301.8236@cc.ruu.nl> Sender: leef@RUUCI12 (Bas Leeflang) Organization: ACCU References: Date: Thu, 28 Apr 1994 07:53:01 GMT Lines: 68 In article , ltaylor@peptidarus.UWaterloo.ca (Lorne Taylor) writes: |> |> I've assigned all the protons in a 16 and a 24 residue peptide in DMSO. |> Now I want to try and get distance constraints from the NOE volumes. |> So: |> 1) If I want to try to extract buildup curves for the NOE's, |> how do I pick a baseline (below is noise/above is data) that |> is the same for all mixing times. Is the exact baseline value crucial? |> I suspect it is as even small variations in the level could |> cause big volume changes. The point is that you must not define a baseline/baseplane that serves as a tresholt. Every where in your spectrum there is noise, which we may assume to have a zero integral for a selected area. The simplest approach is therefor to just add all the points in a designated area, which you define as the peak. Refinements of this approach are possible that are based on 'smarter' integration algorithms, but in general this will result only in minor changes. In the method I just proposed it is important to do a baseline correction prior to integration. (This ensures a zero average of the integral in a non- peak area. |> |> 2) The resolution is fairly high (2k real X 512 expts) and I've zero |> filled to 2K X 2K. My problem is some peaks are quite well resolved |> and some are not (ie SOME amide to alpha protons show a DQCOSY-like |> square array (AX) and others are a single relatively broad peak). |> Since the well resolved peaks have much less area than unresolved ones |> would not the volumes also be uncomparable?? Don't worry to much about the line width of your peaks. A braod peak will generaly have a lower hight, but its integral may be as big as that of a narrow high intensity peak. You just select an area for integration according to the peak widths. Also the COSY like peaks are not to worrysome. It may look ugly but the integral of this antiphase component in your peak will be approximately zero and will therfore not influence the outcome of the integration. |> |> 3) I've some qualitative idea of the conformation (ie NH-NH (i,i+1)) |> NOE in the presence of stronger (NH-alpha H (i,i+1)) which I interpret |> as a peptide that spends some time as a helix but more time in |> unordered conformations. If I put mutually exclusive contraints |> such as these into a distance geometry or restrained MD program, |> would I get a meaningful answer??? Distance geometry and restrained MD are generaly focussed on one single conformation. However, techniques exist that use time-averaged and/or ensemble averaged constraints, that allow several conformations to exist (I do not have any references at hand, so who does?????) Or you could use a program that I wrote a few years ago named CROSREL. This program is not advanced in the sence that it generates structures, but if you know from e.g. MD calculations that a few different conformations are likely to be present Crosrel can combine those models and determine which linear combination gives the best reconstruction of the experimental NOESY or ROESY spectra. (J. Biomol NMR 2 (1992) 495-518, anon. ftp to ruucj1.chem.ruu.nl directory /pub/crosrel) I hope this satisfies your questions, Bas Leeflang ---------------------------------------------------------------------- Bas R. Leeflang _/ _/ _/ _/ _/_/_/ Department of the Bijvoet Center _/_/ _/ _/_/ _/_/ _/ _/ for Biomolecular Research _/ _/ _/ _/ _/ _/ _/_/_/ Padualaan 8, 3584 CH Utrecht _/ _/_/ _/ _/ _/ _/ Tel. : int+31.30.533295 _/ _/ _/ _/ _/ _/ Fax : int+31.30.537623 Email: leef@nmr.chem.ruu.nl From owner-structural-nmr@net.bio.net Thu Apr 28 23:00:00 1994 Newsgroups: bionet.structural-nmr Path: biosci!daresbury!trane.uninett.no!sunic!news.lth.se!news.lu.se!heimdall!bryan From: bryan@fkem2.lth.se (Bryan Finn,,) Subject: Re: NMR data treatment: wrong problem? Message-ID: <1994Apr29.143826.20140@nomina.lu.se> Sender: news@nomina.lu.se (USENET News System) Nntp-Posting-Host: heimdall.fkem2.lth.se Reply-To: bryan@fkem2.lth.se Organization: Physical Chemistry 2, Lund University, Sweden References: <2pqse3$5tg@mserv1.dl.ac.uk> Date: Fri, 29 Apr 1994 14:38:26 GMT Lines: 42 In his recent posting, Per Kraulis writes: >Problem 1: volume integration >... >Problem 2: assignments >... >Problem 3: accounting for the cross peaks >... etc. These are certainly problems to be addressed, and certainly somone or a number of people has thought about it, besides Per of course. Or is that naive to assume so? With the advent of automatic assignment or assisted assignment (which apparently generated a lot of noise at the recent ENC), there must be some sort of semi-objective rules for assessing the correctness of the assignment built in to these programs, aren't there? With respect to the volume integration, this is perhaps less potentially disasterous than misassignment. However, I have noticed a recent trend in some groups to use the peak heights, rather than the volumes, when calculating NOE buildups (or relaxation curves). Is there an underlying reason for this? Do the resulting curves have a better correlation than the corresponding curves generated from peak volumes? Bryan ___________________________________________________________________ | | | Dr. Bryan Finn | | Department of Physical Chemistry 2 Tel: +46-46-108254 | | Chemical Center Fax: +46-46-104543 | | University of Lund | | POB 124 | | S-221 00 Lund Sweden e-mail: bryan@freja.fkem2.lth.se | |_________________________________________________________________| |