=+= README =+============Last update: April 4, 2000 ============
the latest version of this document can be found at:
ftp://ftp.ncbi.nih.gov/fa2htgs/README
-----
After having consulted with NCBI staff (see contact information below)
submitters from Genome Sequencing centers will establish what the best
protocol will be for them to deposit their sequence submission data to
NCBI.
One of these protocol may require the fa2htgs tool, present in this
directory. fa2htgs is a program used to generate Seq-submits (an ASN.1
sequence submission file) for high throughput genome sequencing
projects. Presently we have built fa2htgs for the following platforms:
alphaOSF1.tar.Z
ibmaix.tar.Z
linux.tar.Z
sgi.tar.Z
solaris.tar.Z
sun.tar.Z
win32/fa2htgs.exe (win95/NT)
If fa2htgs is required for a platform not present here,
please let us know (address below) and we will be happy to
try to provide it.
fa2htgs will read a FASTA file (or an Ace Contig file with Phrap sequence
quality values), a Sequin submission template file, (to get contact
and citation information for the submission), and a series of command line
arguments (see below). This program will then combines these
information to make a submission suitable for GenBank. Once you have
generated your submission file, you need to follow the submission
protocol (see the README present on your FTP account or mailed out to
your Center).
fa2htgs is intended for the automation by scripts for bulk submission of
unannotated genome sequence. It can easily be extended from its current
simple form to allow more complicated processing. A submission
prepared with fa2htgs can also be read into Sequin, and then annotated
more extensively. See the Sequin home page at:
http://www.ncbi.nlm.nih.gov/Sequin/
. Contacting NCBI about HTGS submissions and about using fa2htgs:
Questions and concerns about this processing protocol, or how to
use this tool should be forwarded to:
htgs@ncbi.nlm.nih.gov.
=========+=========
using fa2htgs:
typing "fa2htgs -" will cause the program to show its command line
arguments. Below we show these with additional comments (what we show
within { } does not appear on the command line)
fa2htgs 2.0 arguments:
-i Filename for fasta input [File In]
default = stdin
-t Filename for Seq-submit template [File In]
default = template.sub
-o Filename for asn.1 output [File Out] Optional
default = stdout
-e Log errors to file named: [File Out] Optional
-n Organism name? [String] Optional
default = Homo sapiens
-s Sequence name? [String]
{ The sequence must have a name that is unique within }
{ the genome center. We use the combination of the genome }
{ center name (-g argument) and the sequence name (-s) to }
{ track this sequence and to talk to you about it. }
{ The name can have any form you like but must be unique }
{ within your center.
-l length of sequence in bp? [Integer]
{ The length is checked against the actual number of }
{ bases we get. For phase 1 and 2 sequence it is also }
{ used to estimate gap lengths. For phase 1 and 2 }
{ records, it is important to use a number GREATER than }
{ the amount of provided nucleotide, otherwise this will }
{ generate false 'gaps'. Here is assumed that the }
{ putative full length of the BAC or cosmid will be used. }
{ There should be at least 20 to 30 'n' in between the }
{ segments (you can check for these in Sequin), as this }
{ will ensure proper behavior when this sequence }
{ is used with BLAST. Otherwise 'artifactual' unrelated }
{ segment neighbors may be brought into proximity of }
{ each other. }
-g Genome Center tag? [String]
{ This is probably the same as your login name on the }
{ NCBI FTP server }
-p HTGS phase? [Integer]
default = 1
range from 1 to 3
{ Phase 1 - a collection of unordered contigues with }
{ gaps of unknown length. Phase 1 record must }
{ at the very least have two segments with }
{ one gap. }
{ Phase 2 - a series of ordered contigs, gap lengths may }
{ be known. This could be a single sequence, }
{ without gaps, if the sequence has ambiguities }
{ which will be resolved. }
{ Phase 3 - a single contiguous sequence. This sequenced }
{ is finished, although it may, or may not }
{ be annotated. }
-a GenBank accession (if an update) [String] Optional
{ this argument is required if this is an update, do }
{ not use it if you are preparing a new submission }
-r Remark for update? [String] Optional
{ if this is an update, you can add a brief comment }
{ (within "") describing the nature of the update }
{ ("new sequence", "new citation", "updated features") }
-c Clone name? [String] Optional
{ will appear as /clone in the source feature }
{ This could be the same as the -s argument (sequence }
{ name) but this one will appear in the /clone qualifier }
-h Chromosome? [String] Optional
{ will appear as /chromsome in the source feature }
-d Title for sequence? [String] Optional
{ the text that will appear in the DEFINITION line }
{ of the GenBank flatfile. }
-m Take comment from template ? [T/F] Optional
default = F
-u Take biosource from template ? [T/F] Optional
default = F
-x Secondary accession number, separate by commas if multiple, s.t. U10000,L11000 [String] Optional
[ ACCESSION AC000000 L00000 }
{ ^ ^ }
{ | secondary accession number }
{ primary accession number }
{ }
{ In some cases a large segment will supercede another }
{ or group of other accession numbers (records). These }
{ records which are no longer wanted in GenBank should be }
{ made secondary. Using the -x argument you can list the }
{ Accession Numbers you want to make secondary. This will}
{ instruct us to remove the accession number(s) from }
{ GenBank, and will no longuer be part of the GenBank }
{ release. They will nonetheless be available from Entrez.}
{ }
{ !!GREAT CARE should be taken when using this argument!!!}
{ inproper use of accession numbers here will result in }
{ the innapropriate withdrawal of GenBank records from }
{ GenBank, EMBL and DDBJ. We provide this parameter as }
{ a conveniance to submitting centers, but this may need }
{ removed if it is not used carefully. }
-C Clone library name? [String] Optional
{ will appear as /clone-lib="string" on the source feature }
-M Map? [String] Optional
{ will appear as /map="string" on the source feature }
-O Filename for the comment: [File In] Optional
{ will read the comment from a given file. }
{ maximum 100 characters per line. }
{ new lines can be incorporated with "~", and if you }
{ actually want to include the "~" in your text, you }
{ need to escape it with "`". Please ensure that the }
{ correct format is obtained by viewing your comment }
{ in Sequin. }
-T Filename for phrap input [File In] Optional
{ Using this argument infers that you are NOT using the }
{ -i above }
-P Contigs to use, separate by commas if multiple [String] Optional
{ if -P is not indicated with the -T option, then the }
{ fragments will go in in the order that they are in the }
{ ace file (which is appropriate for a phase 1 record, }
{ but not for a phase 2 or 3. If you need to set the }
{ order of the segments of the ace file, you need to set }
{ it with the -P flag, like this: }
{ -P "Contig1,Contig4,Contig3,Contig2,Contig5" }
-A Filename for accession list input [File In] Optional
{ Using this argument infers that you are NOT using the }
{ -i or -T arguments above. The input file contains a }
{ tab-delimited table with three to five columns, which }
{ are accession number, start position, stop position, }
{ and (optionally) length and strand. If start > stop, }
{ the minus strand on the referenced accession is used. }
{ A gap is indicated by the word "gap" instead of an }
{ accession, 0 for the start and stop positions, and a }
{ number for the length. }
-X Coordinates are on the resulting sequence ? [T/F] Optional
default = F
{ if -X is TRUE, then the coordinates in the input file }
{ are on the resulting segmented sequence. This implies }
{ that bases 1 through n of each accession are used. }
{ if -X is FALSE, the coordinates are on the individual }
{ accessions, and these need not start at base 1 of the }
{ record. }
-D HTGS_DRAFT sequence? [T/F] Optional
default = F
-S Strain name? [String] Optional
-b Gap length [Integer]
default = 100
range from 0 to 1000000000
-N Annotate assembly_fragments [T/F] Optional
default = F
-6 SP6 clone (e.g., Contig1,left) [String] Optional
-7 T6 clone (e.g., Contig2,right) [String] Optional
-L Filename for phrap contig order [File In] Optional
{ This is a tab-delimited file that can be used to drive }
{ the order of contigs (normally specified by -P), as well }
{ as indicating the SP6 and T7 ends. It can also be used }
{ when contigs are known to be in opposite orientation. }
{ For example: }
{ }
{ Contig2 + 1 SP6 left }
{ Contig3 + 1 }
{ Contig1 - T7 right }
{ }
{ The first column is the contig name, the second is the }
{ orientation, the third is the fragment_group, the fourth }
{ indicates the SP6 or T7 end, and the fifth says which }
{ side of SP6 or T7 end had vector removed. }
Presented here is an example of a phase 2 submission from an Arabidopsis
sequencing center. It is followed by an command line arguments used in
an example with a Phrap ace file.
BEFORE YOU BEGIN: fa2htgs does depend on the presence of some external
files. These are provided with Sequin, so if a networked version of
Sequin is already installed (see URL above for Sequin info) all the
default files that need to be present will be there and allow fa2htgs
to run.
Here are the files you need (let's assume we have a 100Kb BAC):
1) fasta file (example below)
2) sequin submission file (more on this below)
3) genome center name ("pgec" in this example, use your
FTP login name)
4) the sequence/clone name (this will *always* stay with the record)
5) The phase number:
phase 1: multiple pieces, not in order (alway >= 2 pieces,
often many more)
phase 2: multiple pieces, in order, but can be as few as
one unfinished sequence
phase 3: 1 piece, where the sequence is "finished"
6) the full sequence length, when the project is finished (eg 100000
in our example).
7) A new submission has no Accession Number, and and an update always
does. You will need to keep track of this (ie which sequence name has
which accession number)
8) The organism, in this example "Arabidopsis thaliana"
9) The chromosome number, 1 in this example.
10) the output (file name) convention so far has been to call it the
clone name.ss (eg P74A8.ss) "ss" is a seq-submit, or sequence
submission. We then have our scripts/code report with the same file
name convention. Also note that because we are working in Unix space,
'case' of letter is important, and try to avoid 'metacharacters'
(like ^*/\ etc).
so the phase 1 or 2 FASTA file will look like this (in this example,
this is one has 3 segments, but you could (in phase 1) have many more):
>P74A8 pcr product joining p130c12 and p91c10
gatcagcccaaagcattgattaggggaacttacctgtagagggctgcagcaatggggaac
acctggctgggtcacagagtggtcaatgcactccatgacttttgggtcaggacacagaaa
gaaagagcggggaaccggggggccctacagtgatgaattatactaactgattttagaatg
>?
>fake next line
ttaaacaaacattgcatttccagaataaaccccatttagtaacgcatagtgtgcttgtat
ctcagcctcccaaagtgctgggattatagacatgagccagcgcacctggctttgttagcc
>?200
>fake another line
ttttcaaataactttttgaactttgttaattttttaattgcacgttttctccttcattta
ctaattccattcaaaagtagcatcaatgagaataaattacttaggaatacatttaattaa
aaagtgctagacttgtacactgaaaattacaaagtactctggagatatattc
The first line has the seqence id, and a title, then each segment
is seperated by
>?
>foobar
or:
>?200
>foobar
where you put a "?" if you don't know the distance between the pieces,
or a number of bp if you do know the distance (eg 200 bp), and the
other line is the fasta formated next segment (foobar). So that is it
for phase 1 or 2. Phase 3 will be a single fasta file. All phase 1
will probably always be >?.
So the other thing you need is a submisssion prepared by sequin. This
will allow you to put in the references, authors, Titles, submission
information the way you want it. You simply need to make a 1 bp
submission really. fa2htgs will read that file and copy the
information over to the htgs information with the "real" data.
So once you have made the submission, you deposit it on the FTP account
under "SEQSUBMIT" directory, we have software that looks for it there
every day, validate the center, clone (sequence) id's, check if it's an
update and so on, and write a report that you can pick up the next
day.
It is good to put the output of fa2htgs in Sequin and validate the
record. This is specially important for phase 3 records where many
annotations may be present (added with the help of Sequin): Sequin has
a very good validation suite (look under Search -> Validate)
This finished record is now ready for deposition to your FTP account
in the SEQSUBMIT directory.
example of the command line arguments using quality score/Phrap ace file
(all on tyhe same command line):
./fa2htgs -t nuc1.sqn -o test.cmd32.out -s Phrap_Contig_Test2 -l 111505
-g pgec -p 2 -h 1 -d Phrap_Contig_Test2 -n "Arabidopsis thaliana"
-T g5129z079.ace -P "Contig1,Contig2,Contig4,Contig3,Contig7"
example of a contig file for a yeast chromosome (with coordinates on the
individual accessions):
U73805 1 2669
U12980 79 103687
L05146 133 29410
L22015 2001 41988
L28920 148 54812
-- Questions about fa2htgs or how to submit?
Just contact us at NCBI:
e-mail: htgs@ncbi.nlm.nih.gov
==============+= end of the fa2htgs README =+==========================
