More on: Primers database / gopher.

BENNY SHOMER 9238 pc386 at
Sun Sep 26 12:19:58 EST 1993

Please disregard the previous message.
I have made up my mind to repost the summary anyway.
This is the original posting from last week:

Dear Fellow Bionetters.

I was sincerely happy to find such an interest and involvement in the creation
of the primers database. Many thanks to all those who have replied so far.
I have tryed to compile the threads + personnal communications into 
a cummulated answer. I have deletted a lot of stuff, including repeating 
suggestions. I would like to exclaim that my answers are only first 
thoughts and not final 
desicions. Things will have to be boiled down with Dan, and given further 
thought. So I basically see the discussion as yet alive, since it was very
constructive so far...
BTW, the replies were saved without the '>' citation marks, 
hence the format...
>.....     {Begining of citation}
....>      { End of citation }

Sincerely yours, Benny.
From: lsimon at (Luc Simon)

why not make each record contain all the information pertaining to a given
PCR assay. In other words, make it a database of PCR assays instead of a
database of PCR primers. This would really only translate in adding all
information about the "matching primer" to the proposed form. This way, the
user of the database will not need to parse a possibly large number of
records to pair matching primers manually. The database would also be more
compact since information such as product lenght and cycle conditions need
not be repeated twice. An important benefit of that approach is that Primer
Names do not need to be unique, as is implied in the suggested form where
the user is expected to be able to fetch the matching primer only knowing
its name.

This is a very good idea. I think it will be applicable. Initially, I thought
of such a possibility and couldn't make up my mind. But now it sounds 

I suggest that enough space is allowed in the Species field to allow proper
description of "multi-purpose assays", using degenerate primers or primers
derived from conserved genes that could be useful on a variety of

The fields length was typed in regardless the real length. This also
applies for other fields as well.

In the cycle conditions, the type (make and model) of thermocycler used
might be specified.

Correct! I simply forgot :-). Sorry...

From: pnh at (Paul N Hengen)

***> Tm of primer and based on which method?

Some problem lyes here, as there is no single common method. This will have
to be left in the authors hands, and may be referenced in the remarks field.
I myself, have found out of my experience, that the WhiteHead's 'Primer'
program calculates the Tm best. (i.e. closer to the real Tm in the field...).

***> Computer program used to select primer?

I myself use something like 5 different programs. I would think that such
data may be left for personnal communications (There's an Email@ field,
remember?...). So if one really wants to fetch such data he/she can write.

***> Reason for selecting primer in this location? (sometimes I like
to include or exclude a natural restriction site for later manipulation
although the primer may not be the "best" match for the amplification)

See my answer to the previous question. I believe it applies here as well.

> **********************   Buffer Constituents  ****************************

***> Other additives used?

Right. Maybe we shall even unite some of these fields into one, and treat
this field as text rather than numeric...

> Make: _____________
***> I assume this is the make and model of the PCR machine?

No, This meant for the make of the Polymerase (Very Very important...!).
I will change the text accordingly, to be more understood.

From: Roland.Gronroos at (Roland Gronroos)

> Primer Name: ______________________________
I would like to see a stanard form here. It could be based on the
sequence name from which the primer was choosen. Example "Ecoli uidA
gene pos 499"

We may be entering into a hornet's nest here. I personally think this should
await a workshop, that would set up standards for nomenclature. Or at least
open a seperate thread in the group discussing this matter... 

From: Keith Hutchison <KEITHH at MAINE.MAINE.EDU>

I would suggest that you include a section for 'touchdown' conditions
as well.

Actually there is no need. We just have to treat the cycle conditions field
as text (that's for us to do...). Then, the annealing step can be wrriten as
'65 dC -0.5 dC X 16 cycles' . This means you start at 65, and touch down to

From: finney at Frodo.MGH.Harvard.EDU (Michael Finney)

It is not exactly clear to me what one would put in the "flanking bases" line.
Does this mean the beginning and end of the product?  With what matching
primer?  Including the primer sequence or not?  And using what sequence
numbering system?  An accession number would be ideal, but failing that, the
numbering should be from some explicit reference point (different for each
gene; perhaps the AUG or the 5'-most transcription start site).

Yes. It means begin/end of product. (assuming it's an assays database...).
Including primer sequence (the primer is actually part of the sequence, unless
an irregular one. Then foreign bases will be excluded.). The numbering system
is in the author's responsibility. Perhaps it needs a separate field?...

Since many primers contain sequence alterations, some description is needed of
which bases are complementary to the target.

That's in 'Remarks'...

A nice feature would be to have the database site do a TM calculation (both
for the whole primer and the target-complementary region?).  This would
guarantee that all TM estimates in the database would be done the same way, so
that users could compare apples to apples.

Since we want a WORKING AND TESTED primers database, the Tm should NOT be
calculated. It should be entered by the author as is!  (s.a. another answer

Under Cycling Conditions, it is important to know amount, source, and type of
target DNA; conditions for amplifying from 1 ng of a clone are different than
for amplifying from 1 ng genomic DNA.

You are right. DNA's method of preparation and quantity is really important.

Under Buffer Constituents, it might be good to list standard conditions and
ask people to note any differences.  This saves time, and doesn't assume
everyone knows what the standard buffer is.

This info can be included in the gopher's directory ReadMe file, as well as
other related standard items, and so it doesn't need to be included with each

From: Klaus.Matthaei at

The idea is a good one but PLEASE have included whether the times given are
Block temperature times or In-tube temperature times and the type of
machine they are using.


From: Michael G. Walker <walker at>

May I suggest that the database include BOTH known working and known
non-working primers. It would be useful to know what has been tried
that failed, and might lead to some understanding of what distinguishes
primers that work from those that don't.

Even thow there is something true in it, I wasn't very excited. I'm afraid
of overloading the database with 'J. of Irreproducible Results' like items.
We might end up having so many trees, we wouldn't be able to see the forest.
I wonder what other people in bionetland think of it. Please respond...

From: Andrew Joseph Kompanek <ak10+ at>

If the locus is polymorphic it seems sensible to include info. about the
polymorphism since PCRable repeated sequence polymorphisms are the mainstay
for linkage analysis.

Fragment/Allele size, Allele frequency and allele designation (if available)
would about cover it.  I suppose you could include an empirical heterozygosity
value as well.  Any other details could be dug out of the literature.

This may be included in 'Remarks', I believe...

Well, That's about it for tonight (9:20pm in Israel now :-)   ). 
Hope I'm doing right.....

Cheers, Benny.

More information about the Methods mailing list