I would suggest depositors take a look at the PDB Exchange
Dictionary and at the following definitions:
_atom_site.auth_seq_id
An alternative identifier for _atom_site.label_seq_id that
may be provided by an author in order to match the
identification
used in the publication that describes the structure.
Note that this is not necessarily a number, that the values do
not have to be positive, and that the value does not have to
correspond to the value of _atom_site.label_seq_id. The value
of _atom_site.label_seq_id is required to be a sequential list
of positive integers.
The author may assign values to _atom_site.auth_seq_id in any
desired way. For instance, the values may be used to relate
this structure to a numbering scheme in a homologous structure,
including sequence gaps or insertion codes. Alternatively, a
scheme may be used for a truncated polymer that maintains the
numbering scheme of the full length polymer. In all cases, the
scheme used here must match the scheme used in the publication
that describes the structure.
_atom_site.label_seq_id
This data item is a pointer to _entity_poly_seq.num in the
ENTITY_POLY_SEQ category.
_entity_poly_seq.num
The value of _entity_poly_seq.num must uniquely and sequentially
identify a record in the ENTITY_POLY_SEQ list.
Note that this item must be a number and that the sequence
numbers must progress in increasing numerical order.
So, at the very least, the PDB's internal database and mmCIF and PDBML
files should be able to handle _both_ the simplified numbering the
annotator wishes to impose, and the more scientifically useful notation
an author might use to place their structure in context. It should be
a "simple" matter of programming for the PDB to produce "PDB" entries done
either way.
One should also note the the entire system of insertion codes does not
make much sense without the broader contextual view of families of
structures.
Regards,
Herbert
At 2:33 PM -0400 9/19/08, Frances C. Bernstein wrote:
I was at the PDB from 1974 - 1998 and closely involved with
processing entries 15 to ~9000. We also designed the "PDB
format". My replies were based on what was done for those 24
years and I cannot address what is currently being done at the PDB.
I do not know if the current PDB staff follows this bulletin
board and I can only suggest that you take this matter up
with the current PDB management, the community, and the PDB
advisory board.
Frances
=====================================================
**** Bernstein + Sons
* * Information Systems Consultants
**** 5 Brewster Lane, Bellport, NY 11713-2803
* * ***
**** * Frances C. Bernstein
* *** [EMAIL PROTECTED]
*** *
* *** 1-631-286-1339 FAX: 1-631-286-1999
=====================================================
On Fri, 19 Sep 2008, Linda Brinen wrote:
I'm actually pleased to read your response and interpretation of
what is allowable and why, Frances. However, it's it pretty stark
contrast to what I was told about 18 months ago when I struggled
(and eventually lost) to preserve a numbering scheme that had a
long standing historical and literature precedence when submitting
a new structure to the PDB.
This was a two-domain protein; the first domain - according to
historical numbering - had a number plus a letter code to indicate
the domain; the second domain, which started again with the number
1 - had no letter code. We were told that that was not allowed. We
wanted to preserve insertions and deletions as well, but were also
strongly discouraged, if not flat out told we could not. While it's
not usually prudent to quote offline e-mail exchanges, I'm going to
snip pertinent pieces of the discussion (I'm leaving the original
spelling errors and text bolding in place) with no indication of
the annotator who wrote these guidelines to our group. Here's part
of one of the many 'exchanges' that was had:
"I understand your point and that certain close research
communities have certain habits and traditions but the PDB serves
to the whole community of structural biology, bioinformatics, to
many educators, students... In all these cases, the simplest
possible numbering of sequences, ideally numbering identical to the
numbering used by the UNP sequence database, is far the most useful
because easiest to understand. I do not say this because it is in
our manuals and help pages but because I have eight years of
experience with annotation of all kinds of structures. I would
therefore very much like to ask you to reconsider the way how you
number your protein, your numbering schema is *interpretation* more
than a mere labeling schema. Needles to say, no sequence numbering
can satisfy this ambition...from my point of view, especially the
jump from 96P back to 1 will cause a lot of confusion and
misunderstanding....look at the problem from a standpoint of a
general naturalist instead of an narrow protease community"
This left us with a mandated 'start from 1 and number sequentially'
format that did exactly the opposite of what you, Frances,
correctly mention as important in any numbering scheme: preserve
relationships with other proteins. We've had to resort to
providing 'translation tables' that identify what people were
expecting to see as numbers for active site residues which now have
new and non-sensical numbering. Is it the end of the world? Of
course not. But neither is it necessarily the best scientific or
logical presentation.
At the risk of inciting a rather....animated...dialogue on this
topic, what has your experience been with this kind of thing (i.e.,
were we just unlucky??) and do current practices make sense and
serve the community??
-Linda
Frances C. Bernstein wrote:
All entries list atoms starting at the N-terminus (or 5') so
connectivity goes in the order of the atoms in the file -
obviously with the possibility of unconnected portions
where the density is inadequate.
The entire philosphy of allowing numbering other than 1 - N
had to do with preserving relationships with other proteins.
The most common use relates to having an initial sequence 1 - N
and then a similar sequence from another species with insertions
and/or gaps. People wanted to be able to talk about the active
site (which was preserved) using the same residue numbers.
Negative numbers came up with additions at the N-terminus.
Offhand, I don't recall why descending numbers were used but
I believe that there is at least one such entry.
Frances
=====================================================
**** Bernstein + Sons
* * Information Systems Consultants
**** 5 Brewster Lane, Bellport, NY 11713-2803
* * ***
**** * Frances C. Bernstein
* *** [EMAIL PROTECTED]
*** *
* *** 1-631-286-1339 FAX: 1-631-286-1999
=====================================================
On Fri, 19 Sep 2008, Ian Tickle wrote:
But what connectivity would be implied by descending numbers: the order
in the file or the order of the numbering? I assume the former,
otherwise what would be the point of having descending numbering? And I
wonder how many programs would baulk at it (or even at ascending
negative numbers?).
-- Ian
-----Original Message-----
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]
On
Behalf Of Frances C. Bernstein
Sent: 19 September 2008 16:44
To: Todd Geders
Cc: [email protected]
Subject: Re: [ccp4bb] Non-sequential residue numbering?
As long as each residue within a chain has a unique identifier
(residue number plus insertion code), there is no restriction
on numbering. The numbers can be in ascending or descending
order, non-sequential, and even negative.
Frances
=====================================================
**** Bernstein + Sons
* * Information Systems Consultants
**** 5 Brewster Lane, Bellport, NY 11713-2803
* * ***
**** * Frances C. Bernstein
* *** [EMAIL PROTECTED]
*** *
* *** 1-631-286-1339 FAX: 1-631-286-1999
=====================================================
On Fri, 19 Sep 2008, Todd Geders wrote:
Hello all,
I have a structure from a non-natural fusion of the truncated
C-terminus
of
one protein with the truncated N-terminus of another. For the
deposition, we
want to keep the numbering as found in the separate proteins. It
looks
something like this:
1 12
| |
....HWVCKDIALLMCFFLEEMSEEP....
| |
754 763
At no point is there an overlap in numbering (i.e. the N-terminal
residue
number is higher than the C-terminal residue number).
Is this numbering scheme supported by the PDB standard? Thus far,
all
of the
software seems to handle it (refmac, Coot, PyMOL, pdb_extract, PDB
precheck &
validation, etc).
Can anyone see a reason to not deposit with this non-sequential
residue
numbering?
~Todd
Disclaimer
This communication is confidential and may contain privileged
information intended solely for the named addressee(s). It may
not be used or disclosed except for the purpose for which it has
been sent. If you are not the intended recipient you must not
review, use, disclose, copy, distribute or take any action in
reliance upon it. If you have received this communication in
error, please notify Astex Therapeutics Ltd by emailing
[EMAIL PROTECTED] and destroy all copies of the
message and any attached documents.
Astex Therapeutics Ltd monitors, controls and protects all its
messaging traffic in compliance with its corporate email policy.
The Company accepts no liability or responsibility for any onward
transmission or use of emails and attachments having left the
Astex Therapeutics domain. Unless expressly stated, opinions in
this message are those of the individual sender and not of Astex
Therapeutics Ltd. The recipient should check this email and any
attachments for the presence of computer viruses. Astex
Therapeutics Ltd accepts no liability for damage caused by any
virus transmitted by this email. E-mail is susceptible to data
corruption, interception, unauthorized amendment, and tampering,
Astex Therapeutics Ltd only send and receive e-mails on the basis
that the Company is not liable for any such alteration or any
consequences thereof.
Astex Therapeutics Ltd., Registered in England at 436 Cambridge
Science Park, Cambridge CB4 0QA under number 3751674
--
Linda S. Brinen
Adjunct Assistant Professor
Dept of Cellular & Molecular Pharmacology and
The Sandler Center for Basic Research in Parasitic Diseases
Phone: 415-514-3426 FAX: 415-502-8193
E-mail: [EMAIL PROTECTED]
QB3/Byers Hall 508C
1700 4th Street
University of California
San Francisco, CA 94158-2550
USPS:
UCSF MC 2550
Byers Hall Room 508
1700 4th Street
San Francisco, CA 94158
--
=====================================================
Herbert J. Bernstein, Professor of Computer Science
Dowling College, Kramer Science Center, KSC 121
Idle Hour Blvd, Oakdale, NY, 11769
+1-631-244-3035
[EMAIL PROTECTED]
=====================================================
