BioBit No 18 (Batch to the Past)

"Robert Harper Finland", CSC HARPER at CSC.FI
Thu Dec 6 17:54:00 EST 1990


     1818181818                       1818181818
     1818181818                       1818181818
     1818  1818   1818   1818181818   1818  1818   1818   181818181818
     18181818            1818181818   18181818            181818181818
     18181818     1818   1818  1818   18181818     1818       1818
     1818  1818   1818   1818  1818   1818  1818   1818       1818
     1818181818   1818   1818181818   1818181818   1818       1818
     1818181818   1818   1818181818   1818181818   1818       1818

                                 No 18
    %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
                          BIO-NAUT NEWSLETTER
                     << EDITED BY ROBERT HARPER >>
    %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
  
   Quite recently on the INFO-GCG LIST at the University of Toronto there
   was some discussion as how to recover old messages from the notebooks.
   Well I wrote about two different ways that you can retreive old
   messages. Many people found the instructions helpful and it has been
   suggested that I clump the whole thing together and make BioBit #18
   out of it... so here goes.

   Way back in BioBit #12 I explained how to use LDBASE to retreive
   individual messages from LISTSERV. Unfortunately LDBASE can only be
   used by folks on BITNET who have the capability to do "interactive"
   messaging... so this rules out ALL of the people in the UK and people
   who are ONLY on Internet sites. However there are a couple of ways to
   get the job done simply by relying on mail. 

   I will cover the two methods in this edition of BioBit.

   ***** Method #1

   By far the easiest way to recover past information is to ask for the
   old notebooks that are stored on LISTSERV. The LISTS at IRLEARN store
   all the messages entered into them in a monthly notebook (archive) which
   contains all the messages distributed by the list during one month.

   For the BioSci LISTS at IRLEARN you would just send an E-mail message to
   LISTSERV at IRLEARN.BITNET You should leave the Subj: line blank and in the
   main body of the mail message put you commands, one per line. The info
   you request will be delivered (in about 24 hrs) to you via e-mail.
   Here are some useful LISTSERV commands:

     REVIEW +BIONEWS
     gets you the names and e-mail addresses of the current
     members in the Bionews list.

     INDEX +BIONEWS
     gets you an index of the notebooks available in the Bionews list.
     They take the form LOGYYMM where YY is the year and MM is the month.

     GET +BIONEWS <filetype>
     gets you that month's notebook e.g. GET +BIONEWS LOG9008
     will get you the notebook for the month of August 1990

   Here is an example of a mail batch file which you could send to 
   LISTSERV at IRLEARN.BITNET

   -------------8< cut here 8<-----------------
   REVIEW BIO+NAUT
   INDEX BIO+JRNL
   GET BIO+JRNL LOG9008
   -------------8< cut here 8<-----------------

   In the immortal words of Bo Diddley "Don't write a cheque your tail
   can't cash." If you are getting LOGS of BIO+JRNL then they will be
   BIG so you should have enough disk space to receive them. As a reminder 
   here is a list of the names of the lists you can query at
   LISTSERV at IRLEARN.BITNET

   	+AGEING		+BIONEWS	BIO+JRNL	+BIOMATR
   	BIO+SOFT	EMBL+DB		+BIOJOBS	+GENBANK
   	GENE+ORG	GNOME+PR	+METHODS	BIO+NAUT
   	SCI+RES		AG+FORST        MOL+EVOL        +PROTEIN

   *********************** IMPORTANT REMINDER ***********************

     Listserv commands for the BIOSCI lists should be addressed only to

                     LISTSERV at IRLEARN.BITNET

             and *NOT* to *ANY OTHER E-MAIL ADDRESS* 
   ******************************************************************

   Well that gets the simple method out of the way. Now lets look at how we
   can be more selective. I will use the examples that I gave in INFO-GCG
   so I do not have to go though all the same proceedures again to generate
   examples. The only thing to remember is that if you want to query the
   BIOSCI lists then you should address you queries to LISTSERV at IRLEARN.BITNET
   instead of LISTSERV at UTORONTO.BITNET


   **** METHOD #2

   These instructions were originally sent to the INFO-GCG list. So they
   are specific for the LISTSERV at UTORONTO.BITNET. If you use your common
   sense you should be able to make the neccessary substitutions and make 
   batch mail files work with the BIOSCI lists at LISTSERV at IRLEARN.BITNET

   ... put on your thinking cap now.

                  << Batch to the Past... Part 1 >>

   You can reterive old messages from INFO-GCG by getting the LOG of the
   notebooks, but what if you want to be more selective? And what if you
   don't have LDBASE to carry on an interactive session with LISTSERV.
   Well Tom Easton recently pointed out that you can still make queries
   by mail using a small BATCH job.

   Here is a rundown on the technique. The process is twofold:
        1) You run a batch job to find the information you
           want from INFO-GCG
        2) You run a second batch job to recover the specific
           items that you found out about in the first batch job.

   It goes like this: All that you need to do is send a batch job to
   LISTSERV at UTORONTO.BITNET and the output of your job will be returned to you
   as DATABASE.OUTPUT.
   So what would be some interesting topics to look for?

SEARCH CLARK AND ALIGN IN INFO-GCG  ! clark cos he writes well... alignment
                                    ! cos it a hot topic.
SEARCH PROSITE IN INFO-GCG          ! cos Amos likes to read the answers
                                    ! before he reads the questions.
SEARCH TUTORIAL IN INFO-GCG         ! everyone has a vested interest somewhere

Here is a simple batch job for you to try... modify it by putting in your
own KEYWORDS and send it off to LISTSERV at UTORONTO.BITNET

--------8<  cut here  8<----------
//Infogcg JOB Echo=YES,Reply-via=MAIL
Database search DD=Mine
//Mine DD *
Search clark and align in info-gcg
index
search prosite in info-gcg
index
search tutorial in info-gcg
index
/*
-------8< cut here  8<-------------

You should get the replies within and hour or so if the net is running
smoothly. The next message will show the results of the above BATCH job.

..... we now wait for the results to arrive...

                      << Batch to the Past... Part II >>

   This is the result of the batch job sent by mail to LISTSERV at UTORONTO
   The turn around time was about 30 minutes from Finland to Canada. The
   information is given in the form of a table.

ITEM #  The specific number of a message in the database
Date    When the message was entered
Time    The time it was entered
Recs    The number of lines in the message (approx)
Subject The subject of the message (Good subjects are essential)

   For example Item 000076 was written by Clark or mentions the name
   Clark, and it had the string "align" in it but the subject gives no
   clue as to what the message was about. So use good subjects in your
   mail messages cos it will certainly help people who want to do searches
   on the datadase.

> Search clark and align in info-gcg
--> Database INFO-GCG, 16 hits.

> index
Item #   Date   Time  Recs   Subject
------   ----   ----  ----   -------
000075 90/01/26 16:12  123   Wish list for version 7
000076 90/01/29 10:25   71
000077 90/01/30 11:33   36   Re: Wish list for version 7
000083 90/02/02 16:19   49   RE:
000093 90/02/05 11:13   40   profiles and additions for version 7 (or before?)
000097 90/02/05 12:41   54   flame
000117 90/02/07 23:00  151   How much HELP do we need ?
000118 90/02/08 00:47   92   RE: How much HELP do we need ?
000180 90/03/05 18:45   22   RE: statistical significance
000200 90/03/16 10:36   89   RE: Sequence Analysis Primer
000267 90/05/25 17:26   34   Version 7: Prosite and more
000325 90/07/17 17:39  426   DCL script for EMBL's FastA mail service
000436 90/11/12 10:11   64   RE: multiple sequence alignment
000438 90/11/13 11:14   67   multiple sequence alignment
000439 90/11/13 22:30   96   Volunteers sought
000443 90/11/15 09:25   45   Multiple sequence alignment

> search prosite in info-gcg
--> Database INFO-GCG, 27 hits.

> index
Item #   Date   Time  Recs   Subject
------   ----   ----  ----   -------
000040 89/11/03 22:28  540   Data banks corss-referencing.
000049 89/11/20 20:11  276   PROSITE Newsletter Issue Nb. 1
000091 90/02/05 10:13   36   RE: Database for Consensus-Sequences?
000093 90/02/05 11:13   40   profiles and additions for version 7 (or before?)
000101 90/02/06 12:29   70   ProSite database for consensus sequences
000106 90/02/06 22:58   98   PROSITE
000109 90/02/06 23:04  109   RE: Prosite
000127 90/02/09 21:59   49   PROSITE in machine readable form
000165 90/02/24 16:03  109   Pattern Search in GCG
000189 90/03/09 17:52   37   Intro, Prosite, and Skiing
000193 90/03/13 13:31   76   RE: SWISS-PROT and GenPept
000203 90/03/25 22:50   72   RE: genpept/swissprot
000260 90/05/09 21:54   76   Databases announcements
000263 90/05/16 14:45   26   PROSITE to FIND conversion program available
000267 90/05/25 17:26   34   Version 7: Prosite and more
000272 90/06/04 23:36   56   Prosite Index
000273 90/06/05 10:46   29   Re: Version 7: Prosite and more
000279 90/06/06 09:27   17   Version 7: Prosite
000282 90/06/06 10:04   31   RE: Version 7: Prosite
000284 90/06/07 14:26   37   improved PROSITE conversion program
000285 90/06/08 16:49   18   Prosite Database
000286 90/06/09 11:33   50   RE: Prosite Database
000287 90/06/10 23:57   22   Prosite Database
000340 90/08/01 12:36   17   new database
000422 90/11/05 11:20   33   LDBASE search on "Magic"
000437 90/11/12 22:09   28   RE: LDBASE search on "Magic"
000442 90/11/12 16:24   42   Programs to search Prosite

> search tutorial in info-gcg
--> Database INFO-GCG, 14 hits.

> index
Item #   Date   Time  Recs   Subject
------   ----   ----  ----   -------
000113 90/02/07 10:35   36   RE: The lost users
000117 90/02/07 23:00  151   How much HELP do we need ?
000130 90/02/12 15:14   68   Re: User Interface
000229 90/04/05 14:06   34   GCG tutorials-is anyone happy with theirs?
000303 90/07/08 20:39   39   GCG tutorial.
000349 90/08/08 12:36   39   Tutorial for GCG (Joint writing working group)
000355 90/08/21 18:06  131   How to teach GCG
000362 90/08/22 18:09  274   Biocomputing Tutorial by Reinhard Doelz
000363 90/08/23 13:17   35   FTP preview of BIOLOGICAL TUTORIAL.
000364 90/08/23 14:26   58   Biocomputing Manual Preview
000367 90/08/23 18:34   31   BIOCOMPUTING TUTORIAL.
000389 90/09/23 05:38   27   BIOCOMPUTING TUTORIAL
000390 90/09/24 09:24   21   problems with zoo'ing manuals
000418 90/10/30 17:50   47   Biocomputing tutorial.

   Once you have got the indexes that you are interested in you now have
   to ran another batch job to to recover them. I will deal with that next.


                       <Batch to the Past... Part III >>

   From the first mail batch job we got an index and ITEM # 000076 written
   on the 29/01/90 did not have any subject, and ITEM # 000083 was a reply
   to  this message with no name... so as part of the exercise let's
   recover mystery message ITEM # 000083.... to see what it was, and anyways
   its shorter:-)

****** EXTRACT FROM FIRST BATCH JOB *******
> index
Item #   Date   Time  Recs   Subject
------   ----   ----  ----   -------
000076 90/01/29 10:25   71
000083 90/02/02 16:19   49   RE:

   We now repeat the process. We send a mail batch file to
   LISTSERV at UTORONTO.BITNET See how I keep repeating the LISTSERV
   address get that right... always send to LISTSERV and not to the 
   LIST... (no matter how many times this is repeated someone always gets 
   it wrong). Leave the Subj: line blank and in the main body of the text 
   put the following batch job.

------------8< cut here 8<--------------
//Infogcg JOB Echo=YES,Reply-via=MAIL
Database search DD=Mine
//Mine DD *
Search clark and align in info-gcg
index
print 000083
/*
-----------8< cut here 8<---------------

And in the fullness of time listserv sends back the DATABASE.OUTPUT
and we can discover the secrets of the mystery message with no name.

   DATABASE.OUTPUT
   > Search clark and align in info-gcg    ! SEARCH PART of BATCH JOB
   --> Database INFO-GCG, 18 hits.

   > index                                 ! INDEX PART of BATCH JOB
   Item #   Date   Time  Recs   Subject
   ------   ----   ----  ----   -------
   *** Rest of the list deleted  ! 448 and 449 have clark and align
                                 ! in the body of the message so
                                 ! so they are now added to the
                                 ! index bringing the total to 18
   000448 90/11/16 09:14   52   << Batch to the past... Part I >>
   000449 90/11/16 09:37  106   << Batch to the Past... Part II >>

   > print 000083                           ! PRINT PART of BATCH JOB
   >>> Item number 83, dated 90/02/02 16:19:10 -- ALL

   Lisa Caballero  (Genetics at Salk) writes:

   /     Regarding Stephen Clarks message, I have a few comments:
   /                       ^^^^^
               ***  HERE IS KEYWORD CLARK ****
   /       [stuff deleted]
   /for a contig of lenth 40, its much too stringent when aligning
                                                          ^^^^^
                                         *** HERE IS KEYWORD ALIGN ****

   ****** Epilogue

   Well that is about it. Back to the Future only had three parts, and
   Batch to the Past only has three parts. So I hope this opens up a new
   avenue to explore the backlogs on the BIOSCI lists at 
   LISTSERV at IRLEARN.BITNET

   Happy experimentation.

   -=ROB=-




More information about the Bioforum mailing list