Data Formats solved by 2865

Same Data, Different Formats

Figure 1. An example of a GenBank record header.

A number of different data presentation formats have been used to represent genetic strings. The history of file formats presents its own kind of evolution: some formats have died out, being replaced by more successful ones. Three file formats are currently the most popular:

• FASTA (.fas, .fasta): used by practically all modern software and databases, including ClustalX, Paup, HyPhy, Rdp, and Dambe.
• NEXUS (.nex, .nexus, .nxs): used by Paup, MrBayes, FigTree, and SplitsTree.
• PHYLIP, or "Phylogeny Inference Package" (.phy): used by Phylip, Tree-Puzzle, PhyML, and a number of databases.

A simple reference on file formats can be found here.

In this problem, we will familarize ourselves with FASTA. We will save the other two formats for later problems.

In FASTA format, a string is introduced by a line that begins with '>', followed by some information labeling the string. Subsequent lines contain the string itself; the next line beginning with '>' indicates that the current string is complete and begins the label of the next string in the file.

GenBank hosts its own file format for storing genome data, containing a large amount of information about each interval of DNA. The GenBank file describes the interval's source, taxonomic position, authors, and features (see Figure 1).

A sample GenBank entry can be found here. You may export an entry to a variety of file formats by selecting the appropriate file format under the Send To: dropdown menu at the top of the page.

Problem

GenBank can be accessed here. A detailed description of the GenBank format can be found here. A tool, from the SMS 2 package, for converting GenBank to FASTA can be found here.

Given: A collection of $n$ ($n \leq 10$) GenBank entry IDs.

Return: The shortest of the strings associated with the IDs in FASTA format.

Sample Dataset

FJ817486 JX069768 JX469983

Sample Output

>JX469983.1 Zea mays subsp. mays clone UT3343 G2-like transcription factor mRNA, partial cds
ATGATGTATCATGCGAAGAATTTTTCTGTGCCCTTTGCTCCGCAGAGGGCACAGGATAATGAGCATGCAA
GTAATATTGGAGGTATTGGTGGACCCAACATAAGCAACCCTGCTAATCCTGTAGGAAGTGGGAAACAACG
GCTACGGTGGACATCGGATCTTCATAATCGCTTTGTGGATGCCATCGCCCAGCTTGGTGGACCAGACAGA
GCTACACCTAAAGGGGTTCTCACTGTGATGGGTGTACCAGGGATCACAATTTATCATGTGAAGAGCCATC
TGCAGAAGTATCGCCTTGCAAAGTATATACCCGACTCTCCTGCTGAAGGTTCCAAGGACGAAAAGAAAGA
TTCGAGTGATTCCCTCTCGAACACGGATTCGGCACCAGGATTGCAAATCAATGAGGCACTAAAGATGCAA
ATGGAGGTTCAGAAGCGACTACATGAGCAACTCGAGGTTCAAAGACAACTGCAACTAAGAATTGAAGCAC
AAGGAAGATACTTGCAGATGATCATTGAGGAGCAACAAAAGCTTGGTGGATCAATTAAGGCTTCTGAGGA
TCAGAAGCTTTCTGATTCACCTCCAAGCTTAGATGACTACCCAGAGAGCATGCAACCTTCTCCCAAGAAA
CCAAGGATAGACGCATTATCACCAGATTCAGAGCGCGATACAACACAACCTGAATTCGAATCCCATTTGA
TCGGTCCGTGGGATCACGGCATTGCATTCCCAGTGGAGGAGTTCAAAGCAGGCCCTGCTATGAGCAAGTC
A

Programming Shortcut

Here we can again use the Bio.Entrez module introduced in “GenBank Introduction”. To search for particular access IDs, you can use the function Bio.Entrez.efetch(db, rettype), which takes two parameters: the db parameter takes the database to search, and the rettype parameter takes the data format to be returned. For example, we use "nucleotide" (or "nuccore") as the db parameter for Genbank and "fasta" as the rettype parameter for FASTA format.

The following code illustrates efetch() in action. It obtains plain text records in FASTA format from NCBI's [Nucleotide] database.

>>>from Bio import Entrez
>>>Entrez.email = "your_name@your_mail_server.com"
>>>handle = Entrez.efetch(db="nucleotide", id=["FJ817486, JX069768, JX469983"], rettype="fasta")
>>>records = handle.read()
>>>print records

To work with FASTA format, we can use the Bio.SeqIO module, which provides an interface to input and output methods for different file formats. One of its main functions is Bio.SeqIO.parse(), which takes a handle and format name as parameters and returns entries as SeqRecords.

>>>from Bio import Entrez
>>>from Bio import SeqIO
>>>Entrez.email = "your_name@your_mail_server.com"
>>>handle = Entrez.efetch(db="nucleotide", id=["FJ817486, JX069768, JX469983"], rettype="fasta")
>>>records = list (SeqIO.parse(handle, "fasta")) #we get the list of SeqIO objects in FASTA format
>>>print records[0].id  #first record id
gi|227437129|gb|FJ817486.1|
>>>print len(records[-1].seq)  #length of the last record
771