I have previously discussed the way to download the sequence from NCBI database if you have a list of accession numbers. But what if you have a range of accession numbers (e.g. EF100000…………EF102000). If you search by a single accession number in the NCBI GenBank then you have no problem pulling up a record, but obviously you would not like to do this for thousands of EST records. So what is the easiest way to retrieve all these records when you way provide a range of accession numbers simultaneously from GenBank? Actually there are two way to perform this task. You can use either GenBank’s web interface or can go for command
line option also if you are comfortable with. For command line option you can
use, as usual, PERL script to query the GenBank .
1. GenBank’s web interface
This is the easiest way to download multiple sequences from NCBI GenBank if you have a range of accession numbers. It can be done in few steps:
- Go to the NCBI webpage
- Choose the database (protein, nucleotide, EST,GSS)
- Give the range of accession numbers followed by [accn] tag
For example I have an accession number range EF100000…………EF102000 so my query will be like this EF100000:EF102000[accn]. Click HERE for actual example.
- 5 Server for Batch BLAST against NCBI database
- Some Easy Ways to Download Multiple Sequences from NCBI
- Perl Script 6: How to search NCBI database using PERL script
2. Command line option
If you don't like to leave your desktop then command line is made for you. Download the NCBI search PERL script and search the database rite from you shell.
| Script name | Download |
|---|---|
| NCBI search.pl |  |
Uses
perl ncbi_search.pl -q EF100000:EF102000[accn] -o results.txt -d nucleotide -r fasta -m 2000
In this command PERL script will search the nucleotide database with query EF100000:EF102000[accn] and save all 2000 sequences in result.txt file in FASTA format.
Options
-q [STRING] : raw query text (Required)
-o [FILE] : output file to create (Required).
-d [STRING] : name of the NCBI database to search, such as 'nucleotide', 'pubmed' (Required).
-r [STRING] : the type of information requested. For sequences, 'fasta' is often used.
-m [INTEGER] : the maximum number of records to return (Optional)
Post a Comment