GenBank Release 138.0 Now Available
cavanaug at ncbi.nlm.nih.gov
Fri Oct 24 18:59:32 EST 2003
Greetings GenBank Users,
GenBank Release 138.0 is now available via ftp from the National
Center for Biotechnology Information (NCBI):
Ftp Site Directory Contents
---------------- --------- ---------------------------------------
ftp.ncbi.nih.gov genbank GenBank Release 138.0 flatfiles
ncbi-asn1 ASN.1 data used to create Release 138.0
Uncompressed, the Release 138.0 flatfiles require approximately 118 GB
(sequence files only) or 131 GB (including the 'short directory' and
'index' files). The ASN.1 version requires approximately 105 GB. From
the release notes:
Release Date Base Pairs Entries
137 Aug 2003 33865022251 27213748
138 Oct 2003 35599621471 29819397
Close-of-data was 10/19/2003. Five days were required to prepare this
release. In the nine week period between the close dates for GenBank
releases 137.0 and 138.0, GenBank grew by 1,734,599,220 basepairs and by
2,605,649 sequence records. During that same period, 146,488 records
were updated. Combined, this yields an average of about 49,000
new/updated records per day.
We would like to remind our users that GenBank mirrors are available
at ftp://genbank.sdsc.edu/pub and ftp://bio-mirror.net/biomirror/genbank.
Those who experience slow FTP transfers of large files might realize an
improvement in transfer rates from these alternate sites when traffic at
the NCBI is heavy.
For additional release information, see the README files in either of
the directories mentioned above, and the release notes (gbrel.txt) in
the genbank directory. Sections 1.3 and 1.4 of the release notes
(Changes in Release 138.0 and Upcoming Changes) have been appended
*NOTE* Section 1.4.1 discusses a very important change : the removal
of sequence length limits for all classes of GenBank sequence records,
as of June 2004. We strongly encourage all users to review this
Release 138.0 data, and subsequent updates, are available now via
NCBI's Entrez and Blast services.
If you encounter problems while ftp'ing or uncompressing Release
138.0, please send email outlining your difficulties to
info at ncbi.nlm.nih.gov .
Mark Cavanaugh, Vladimir Alekseyev, Anton Butanaev, Michael Kimelman
1.3 Important Changes in Release 138.0
1.3.1 Organizational changes
The total number of sequence data files increased by 29 with this release:
- the EST division is now comprised of 279 files (+9)
- the GSS division is now comprised of 102 files (+18)
- the ROD division is now comprised of 10 files (+1)
- the VRT division is now comprised of 4 files (+1)
Updates to a significant number of HTC sequences has resulted in a
*decrease* in the overall number of HTC sequence files, from four
to three. The changes reduced the sizes of the records, such that
they now fit into only three files.
1.3.2 BASECOUNT line dropped
The BASECOUNT line of the GenBank flatfile format provides totals for
the number of A, T, G, C, and 'other' basepairs that are present within
the sequence of a database record. For example:
LOCUS AY244763 5686 bp DNA linear BCT 10-APR-2003
DEFINITION Rhodococcus sp. DS7 cysDNCQ operon, complete sequence.
VERSION AY244763.1 GI:29725657
BASE COUNT 1137 a 1661 c 1821 g 1066 t 1 others
1 cgcggtttgt gacgtctgat tgccggtcat tgacctttgg gtagaacgag ttctattctg
61 tgattgcgtt caatttagaa ccagtccggt acataaatgt accgatgcgg aaatggtgtt
5281 tgtcagctcg gtgtctggng gcgaggctaa gcaccaacgg cttcggtagc agaaccacat
This information is computationally expensive to produce for very large
sequences, and for sequences in the CON division. In the CON division case,
a record might be comprised of 'pointers' to hundreds, or even thousands,
of underlying GenBank records. So to calculate the BASECOUNT line content,
retrievals of sequence data for those many records must be performed. This
can noticeably impact the response time for flatfile generation within the
As of this October 2003 GenBank Release (138.0), the BASECOUNT linetype
is no longer present in GenBank Release and GenBank Update products.
Depending on demand, a display option might be implemented in Entrez which
allows users to choose to have BASECOUNT shown.
1.3.3 New oriT feature
As of this October 2003 Release 138.0, a new feature key (oriT) is be
legal for the feature table:
Feature Key oriT
Definition origin of transfer; region of a DNA molecule
where transfer is
initiated during the process of conjugation or
Mandatory qualifiers: None
Optional Qualifiers: /bound_moiety="text"
/locus_tag="text" (single token)
Molecule Scope: DNA
Comments: rep_origin should be used for origins of
has legal values RIGHT, LEFT and BOTH, however
only RIGHT and LEFT
are valid when used in conjunction with the oriT
of transfer can be present in the chromosome;
plasmids can contain
multiple origins of transfer
1.3.4 New /ecotype qualifier
As of this October 2003 GenBank Release 138.0, a new source feature
qualifier called /ecotype is legal. The definition for /ecotype is:
Definition A distinct population of organisms of a
widespread species that has adapted
genetically to its own local habitat.
Nevertheless, they can still reproduce
with members of other ecotypes of the
Value format "text"
Comment 'Ecotype' is often applied to standard
genetic stocks of Arabidopsis thaliana,
but it can be applied to any organism,
especially sessile organisms like plants.
1.3.5 Change to value format of /rpt_unit
As of this October 2003 Release 138.0, the value-format of the /rpt_unit
qualifier has been changed to allow 'text' . The previously documented
Value format <feature_label> or <base_range>
Comment used to indicate feature which defines (or base range
repeat unit of which a repeat region is made
However, a very common value for /rpt_unit is a literal sequence
string that represents the repeating unit(s). For example:
So the format of this qualifier has been changed to:
Value format "text" or <feature_label> or <base_range>
Existing feature label and base range values are now presented as
1.3.6 New operon feature
Starting with this October 2003 Release 138.0, a new type of feature
is legal for the feature table:
Feature Key: operon
Definition: region containing polycistronic transcripts
and regulatory sequences containing genes
that encode enzymes that are in the same
Mandatory qualifiers: /operon
Optional qualifiers: /allele="text"
/locus_tag="text" (single token)
In bacteria, many genes encoding for specific biosynthetic pathways are
transcribed in polycistronic operons. It has been challenging to reflect
this biology within the GenBank flatfile format. GenBank has been using
a gene feature that spans the entire regulatory region and all of the coding
regions and then gene features corresponding to the individual genes spaning
the coding genes.
The new operon feature will simplify the annotation of cases like these.
1.3.7 [er] prefix for JOURNAL line
As of this October 2003 Release 138.0, a new prefix is legal for the
JOURNAL line: [er] .
This prefix is an abbreviation for Electronic Resource, which is a
term that describes journal articles that are available on-line.
In 1999, an interim 'Online Publication' REFERENCE format was adopted for
use at GenBank in order to cite articles appearing only electronically:
REFERENCE 1 (bases 1 to 2858)
AUTHORS Smith, J.
TITLE Cloning and expression of a phospholipase gene
JOURNAL Online Publication
REMARK Online-Journal-name; Article Identifier; URL
In subsequent years, no standards for citing on-line journal articles
have emerged from library organizations.
One such library organization (National Library of Medicine, NIH) is now
assigning identifiers (Medline UIs and PubMed Ids) to articles published
on-line, and it is presenting these articles in a manner that is identical
to print-journal articles. For example:
AUTHORS Haas,B.J., Volfovsky,N., Town,C.D., Troukhan,M.,
Feldmann,K.A., Flavell,R.B., White,O. and Salzberg,S.L.
TITLE Full-length messenger RNA sequences greatly improve genome
JOURNAL Genome Biol. 3 (6), RESEARCH0029 (2002)
Although these citations may contain journal abbreviations, volume numbers,
issue/part/supplement numbers, pages, and year (just like a print-journal
citation), there is no guarantee that the contents of these fields will be
comparable to those of print-journal citations.
In the case above, although the page number is a bit unusual
("RESEARCH0029"), software that processes the JOURNAL line would probably
still be able to parse its contents. But there is also a possibility that
these fields could contain unusual characters (embedded spaces, commas,
parentheses), and possibly even URLs. So the presence of [er] :
JOURNAL [er] Genome Biol. 3 (6), RESEARCH0029 (2002)
will act as a warning (primarily to software) that the contents of the
JOURNAL line might not be as parsable as a print-journal JOURNAL line.
1.3.8 GSS File Header Problem
GSS sequences at GenBank are maintained in one of two different systems,
depending on their origin. One recent change to release processing involves
the parallelization of the dumps from those systems. Because the second dump
(for example) has no prior knowledge of exactly how many GSS files will be
dumped from the first, it doesn't know how to number it's own output files.
There is thus a discrepancy between the filenames and file headers of eleven
GSS flatfiles in Release 138.0. Consider the gbgss74.seq file:
GBGSS1.SEQ Genetic Sequence Data Bank
October 15 2003
NCBI-GenBank Flat File Release 138.0
GSS Sequences (Part 1)
87880 loci, 66457089 bases, from 87880 reported sequences
Here, the filename and part number in the header is "1", though the file
has been renamed as "92" based on the files dumped from the other system.
We will work to resolve this discrepancy in future releases, but the
priority is certainly much lower than many other tasks.
1.4 Upcoming Changes
1.4.1 **Sequence Length Limitation To Be Removed In June 2004**
At the May 2003 collaborative meeting among representatives of GenBank,
EMBL, and DDBJ, it was decided that the 350 kilobase limit on the sequence
length of database records will be removed as of June 2004.
Individual, complete sequences are currently expected to be a maximum
of 350 kbp in length. One major reason for the existence of this limit is
as an aid to users of sequence analysis software, some of which might not
be capable of processing megabase-scale sequences.
However, very significant exceptions to the 350 kbp limit have existed
for several years; Phase 1 (unordered, unoriented) and Phase 2 (ordered,
oriented) high-throughput genomic sequences (HTGS) generated by efforts
such as the Human Genome Project; large dispersed eukaryotic genes with
an intron/exon structure that spans more than 350 kbp; and sequences
which result from assemblies of Whole Genome Shotgun (WGS) project data.
Given these exceptions, and the technological advances which have made
large-scale sequencing practical for an increasing number of researchers,
the collaboration has decided that the 350 kbp limit must be removed.
As of June 2004, the length of database sequences will be limited only
by the natural structures of an organism's genome. For example, a single
record might be used to represent all of human chromosome 1, which is
approximately 245 Mbp in length.
Software developers for some of the larger commercial sequence analysis
packages were recently asked what timeframe would be appropriate for this
change. Answers ranged from "immediately", to "several months", to "one year".
So the one-year timeframe was selected, to provide ample time to implement
changes which megabase-scale sequences may require.
Some sample records with very large sequences have been made available
so that developers can begin to test their software modifications:
Many changes are expected after the removal of the length limit. For
example, complete bacterial genomes (typically on the order of several
megabases) will be re-assembled into single sequence records. The submission
process for such genomes will become much more streamlined, since database
staff will no longer have to split the genomes into pieces. BLAST services
will be enchanced, so that hits reported within very large sequences will
be presented in a meaningful context.
All such changes will be discussed more fully in future release notes,
the NCBI newsletter, and the GenBank newsgroup.
1.4.2 Accession format of WGS records
Whole Genome Shotgun (WGS) sequences utilize an accession number format which
is different from those used for non-WGS GenBank sequences. This format is
referred to as 4 + 2 + 6, and is comprised of:
- a 4-letter WGS project code
- a 2-digit assembly-version number
- a 6 (and sometimes 7) digit sequence number
Because of their unique nature, WGS sequences are kept separate from other
For example, sequences much larger than the current 350 kbp limit can be
generated during the WGS assembly phase. In addition, there is no tracking
of nucleotide sequences from one assembly to the next. So the accessions
are not necessarily related in any way to those of the next assembly:
When a WGS project is completed, it is possible that the submittors may chose
to submit a single finished WGS sequence with a 4 + 2 + 6 accession, at which
point it could appear in the non-WGS portion of GenBank.
Alternately, the submittors might chose to submit the completed genome via
a non-WGS method, in which case a de-novo non-WGS accession would be assigned.
That record would then have one or more 2 + 4 + 6 WGS accessions as
Both scenarios are possible, especially after the 350 kbp sequence length
restriction is lifted. So we felt it was important to alert users that
WGS accessions may eventually be encountered in the non-WGS portion of
of GenBank, as primary or secondary accession numbers.
- GenBank newsgroup see: http://www.bio.net/hypermail/genbankb/
- GENBANKB e-mail: messages sent to genbankb at net.bio.net
- subscribe: e-mail biosci-server at net.bio.net with: subscribe genbankb
- unsub: e-mail biosci-server at net.bio.net with: unsubscribe genbankb
- GenBank on the WWW, see: http://www.ncbi.nlm.nih.gov/Genbank/
- problems with GENBANKB? E-mail moderator: francis at cmmt.ubc.ca
More information about the Genbankb