Convert GenBank records to FASTA; feature annotations and references are dropped.
Pulling the sequence out of a GenBank record
Yes, this page converts GenBank to FASTA. Each record's sequence is written as one FASTA record, and its features, references, and other annotations are dropped. A record that lists only its length or assembly pieces instead of its letters is refused, and no sequence-type question is asked.
Sequence kept, annotations left behind
| Data in your file | In the output | Details |
|---|---|---|
| Sequence letters | Kept | In the sample on this page, the 63-letter sequence is unchanged. |
| Record ID and definition | Kept in the sample | The header line is >EX000001 Example organism example gene, complete cds, which is the ID followed by the definition text. |
| Features | Lost | The sample's one feature is gone from the output. |
| References and other annotations | Lost | The notice under the Convert file button says these are dropped. |
| Molecule type | Not stored | FASTA does not record whether a sequence is DNA, RNA, or protein. |
The sample on this page, a 519-byte text record, converts to one record. If Biopython reports changed data during a conversion, the page lists a short note for each kind of change, with counts, and the download still proceeds.
Records that have no sequence letters
Some records, such as a CONTIG record, give only the sequence length or list the pieces of an assembly instead of the letters. FASTA needs the letters, so such a record is refused; the message suggests downloading the record with its sequence, or choosing GenBank or EMBL output.
Proteins, annotations, and the way back
GenBank coding sequences (CDS) to FASTA turns each CDS feature into one protein record built from its stated translation; the rest of the record is not converted. GenBank to EMBL is available for another flat-file format; each record must state its molecule type and have an identifier without spaces. In the sample on this page, the record's one feature is still present after that conversion.
The reverse route, FASTA to GenBank, is available. It asks whether the sequences are DNA, RNA, or protein, requires names without spaces, and cannot restore the features or references dropped here.
If the file is refused for another reason
If the upload cannot be read, the page says only that it could not be read as GenBank and asks you to check that it is valid; no cause is given. A file over 50 MB is refused with the message "The file is too large. The limit is 50 MB."
Run it with Biopython
For batch work, you can use Biopython's SeqIO.convert function. This pair needs no molecule type argument.
from Bio import SeqIO
count = SeqIO.convert("input.genbank", "genbank",
"output.fasta", "fasta")
print("Converted %i records" % count)
Sample input file: GenBank
Copy this example into a file, or download it, to try a GenBank conversion.
Download sample (sample.gb, 519 bytes)
Samples are small illustrative files. Some come from the Biopython test suite (see the project's sample notes); they are not real study data.