Edit this page on GitHub

The module for multiple sequence alignments, AlignIO

This page describes Bio.AlignIO, a new multiple sequence Alignment Input/Output interface for BioPython 1.46 and later.

In addition to the built in API documentation, there is a whole chapter in the Tutorial on Bio.AlignIO, and although there is some overlap it is well worth reading in addition to this page. There is also the API documentation (which you can read online, or from within Python with the help() command).

Aims

You may already be familiar with the Bio.SeqIO module which deals with files containing one or more sequences represented as SeqRecord objects. The purpose of the SeqIO module is to provide a simple uniform interface to assorted sequence file formats.

Similarly, Bio.AlignIO deals with files containing one or more sequence alignments represented as Alignment objects. Bio.AlignIO uses the same set of functions for input and output as in Bio.SeqIO, and the same names for the file formats supported.

Note that the inclusion of Bio.AlignIO does lead to some duplication or choice in how to deal with some file formats. For example, Bio.AlignIO and Bio.Nexus will both read alignments from NEXUS files - but Bio.NEXUS allows more control and the use of trees.

My vision is that for reading or writing sequence alignments you should try Bio.AlignIO as your first choice. In some cases you may only care about the sequences themselves, in which case try using Bio.SeqIO on the alignment file directly. Unless you have some very specific requirements, I hope this should suffice.

Peter

File Formats

This table lists the file formats that Bio.AlignIO can read and write, with the Biopython version where this was first supported.

The format name is a simple lowercase string, matching the names used in Bio.SeqIO. Where possible we use the same name as BioPerl’s SeqIO and EMBOSS.

Format name Reads Writes Notes
clustal 1.46 1.46 The alignment format of Clustal X and Clustal W.
emboss 1.46 No The EMBOSS simple/pairs alignment format.
fasta 1.46 1.48 This refers to the input file format introduced for Bill Pearson’s FASTA tool, where each record starts with a “>” line. Note that storing more than one alignment in this format is ambiguous. Writing FASTA files with AlignIO failed prior to release 1.48 (Bug 2557).
fasta-m10 1.46 No This refers to the pairwise alignment output from Bill Pearson’s FASTA tools, specifically the machine readable version when the -m 10 command line option is used. The default free format text output from the FASTA tools is not supported.
ig 1.47 No The refers to the IntelliGenetics file format often used for ordinary un-aligned sequences. The tool MASE also appears to use the same file format for alignments, hence its inclusion in this table. See MASE format.
maf 1.69 1.69 Multiple Alignment Format (MAF) produced by Multiz. Used to store whole-genome alignments, such as the 30-way alignments available from the UCSC genome browser.
mauve 1.70 1.70 Mauve’s eXtended Multi-FastA (XMFA) file format
nexus 1.46 1.48 Also known as PAUP format. Uses Bio.Nexus internally. Only one alignment per file is supported.
phylip 1.46 1.46 This is a strict interpretation of the interlaced PHYLIP format which truncates names at 10 characters.
phylip-sequential 1.59 1.59 This is a strict interpretation of the sequential PHYLIP format which truncates names at 10 characters.
phylip-relaxed 1.58 1.58 This is a relaxed interpretation of the PHYLIP format which allows long names.
stockholm 1.46 1.46 Also known as PFAM format, this file format supports rich annotation.
       

In addition, you can store the (gapped) sequences from an alignment in many of the file formats supported by Bio.SeqIO. The most common example of this is storing alignments in the simple fasta format. However, storing more than one alignment in a single such file is ambiguous - and this is not recommended.

Alignment Input

As in Bio.SeqIO, there are two functions for alignment input. These are Bio.AlignIO.read() for when the file contains one and only one alignment, and the more general Bio.AlignIO.parse() when the file may contain multiple separate alignments.

Both these functions have two required arguments, a file handle and a file format. As with Bio.SeqIO, Biopython insists that you explicitly give the expected file format, rather than attempting to guess this based on the filename or contents. The file format is specified as a lower case string, see the table above.

As an example, we’ll look at a PFAM seed alignment for the Fibrinogen gamma chain PF09395 Fib_gamma. At the time of writing, this contained 14 sequences with an alignment length of 77 amino acids, and is shown below in the PFAM or Stockholm format:

# STOCKHOLM 1.0
#=GS Q7ZVG7_BRARE/37-110  AC Q7ZVG7.1
#=GS Q6X871_SCAAQ/1-77    AC Q6X871.1
#=GS O02676_CROCR/1-77    AC O02676.1
#=GS Q6X869_TENEC/1-77    AC Q6X869.1
#=GS FIBG_HUMAN/40-116    AC P02679.3
#=GS O02689_TAPIN/1-77    AC O02689.1
#=GS O02688_PIG/1-77      AC O02688.1
#=GS O02672_9CETA/1-77    AC O02672.1
#=GS O02682_EQUPR/1-77    AC O02682.1
#=GS Q6X870_CYNVO/1-77    AC Q6X870.1
#=GS FIBG_RAT/40-116      AC P02680.3
#=GS Q6X866_DROAU/1-76    AC Q6X866.1
#=GS O93568_CHICK/40-116  AC O93568.1
#=GS FIBG_XENLA/38-114    AC P17634.1
Q7ZVG7_BRARE/37-110          GFGTYCPTTCGVADYLQRYKPDMDKKLDDMEQDLEEIANLTRGAQDKVVYLK---DSEAQAQKQSPDTYIKKSSNML
Q6X871_SCAAQ/1-77            RFGSYCPTTCGIADFLSTYQATVDKDLQTLEDILSQAENKTMEAKELVKAIQVSYLPEDPARPNRVELATKDSKKMM
O02676_CROCR/1-77            RFGSYCPTTCGIADFLSTYQTGVXNDLRTLEDLLSGIENKTSEAKELIKSIQVSYNPNEPPKPNTIVSATKDSKKMM
Q6X869_TENEC/1-77            RFGSYCPTTCGIADFLSTYQGSIDKDLQTLEDILNQVENKTXEASELIKSIQVSYNPDEPPRPNMIEGATQKSKKML
FIBG_HUMAN/40-116            RFGSYCPTTCGIADFLSTYQTKVDKDLQSLEDILHQVENKTSEVKQLIKAIQLTYNPDESSKPNMIDAATLKSRKML
#=GS FIBG_HUMAN/40-116    DR PDB; 1qvh L;14-45
#=GS FIBG_HUMAN/40-116    DR PDB; 1fza C;88-90
#=GS FIBG_HUMAN/40-116    DR PDB; 1fzb C;88-90
#=GS FIBG_HUMAN/40-116    DR PDB; 1fzb F;88-90
#=GS FIBG_HUMAN/40-116    DR PDB; 1qvh I;14-45
#=GS FIBG_HUMAN/40-116    DR PDB; 1fza F;88-90
#=GR FIBG_HUMAN/40-116    SS CCXCXBXXHHHHHHHHHHHHHHHHHHHHHHHXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX-CC
O02689_TAPIN/1-77            RFGSYCPTTCGIADFLSTYQTXVDKDLQVLEDILNQAENKTSEAKELIKAIQVRYKPDEPTKPGGIDSATRESKKML
O02688_PIG/1-77              RFGSYCPTMCGIAGFLSTYQNTVEKDLQNLEGILHQVENKTSEARELIKAIQISYNPEDLSKPDRIQSATKESKKML
O02672_9CETA/1-77            RFGSYCPTTCGVADFLSNYQTSVDKDLQNLEGILYQVENKTSEARELVKAIQISYNPDEPSKPNNIESATKNSKRMM
O02682_EQUPR/1-77            RFGSYCPTTCGIADFLSNYQTSVDKDLQDFEDILHRAENQTSEAEQLIQAIRTSYNPDEPPKTGRIDAATRESKKMM
Q6X870_CYNVO/1-77            RFGSYCPTTCGIADFLSTYQTKVDEDLQNLEDILYRVENRTSEAKELIKAIQVDYNPGEPPKQSVTEGATQNAKKMV
FIBG_RAT/40-116              RFGSYCPTTCGISDFLNSYQTDVDTDLQTLENILQRAENRTTEAKELIKAIQVYYNPDQPPKPGMIEGATQKSKKMV
Q6X866_DROAU/1-76            RFGSYCPTTCGIADFLNKYQTTIDQDLRHMEETLRDIDNKTAESTLLIQKIQIGQTPDPRPQ-NVIGDVTQKSRKMI
O93568_CHICK/40-116          RFGSYCPTTCGIADFFNKYRLTTDGELLEIEGLLQQATNSTGSIEYLIQHIKTIYPSEKQTLPQSIEQLTQKSKKII
#=GS O93568_CHICK/40-116  DR PDB; 1m1j F;14-90
#=GS O93568_CHICK/40-116  DR PDB; 1m1j C;14-90
#=GR O93568_CHICK/40-116  SS CCEEEEE-CCCCCCCCCCCCCHHHCCCCCHHHHHHHHHHHHHHHCCCCCCHHHHS-SSTT--SS-HHHHHHHHHHHH
FIBG_XENLA/38-114            RFGEYCPTTCGISDFLNRYQENVDTDLQYLENLLTQISNSTSGTTIIVEHLIDSGKKPATSPQTAIDPMTQKSKTCW
#=GC SS_cons                 CCECEEE-CCCCCCCCCCCCCHHHCCCCCHHHHHHHHHHHHHHHCCCCCCHHHHS-SSTT--SS-HHHHHHHHHHCC
#=GC seq_cons                RFGSYCPTTCGIADFLSsYQssVDcDLQsLEsILpplEN+ToEAc-LIKuIQlsYsP--ss+PstI-uATpcSKKMl
//

You will notice that there is plenty of annotation information here, including accession numbers for each sequence and also some PDB database cross-references and secondary structure information for the human and chick fibrinogen proteins.

This file contains a single alignment, so we can use the Bio.AlignIO.read() function to load it in Biopython. Let’s assume you have downloaded this alignment from Sanger, or have copy and pasted the text above, and saved this as a file called PF09395_seed.sth on your computer. Then in python:

from Bio import AlignIO
alignment = AlignIO.read(open("PF09395_seed.sth"), "stockholm")
print("Alignment length %i" % alignment.get_alignment_length())
for record in alignment :
    print(record.seq + " " + record.id)

That should give:

Alignment length 77
GFGTYCPTTCGVADYLQRYKPDMDKKLDDMEQDLEEIANLTRGAQDKVVYLK---DSEAQAQKQSPDTYIKKSSNML Q7ZVG7_BRARE/37-110
RFGSYCPTTCGIADFLSTYQATVDKDLQTLEDILSQAENKTMEAKELVKAIQVSYLPEDPARPNRVELATKDSKKMM Q6X871_SCAAQ/1-77
RFGSYCPTTCGIADFLSTYQTGVXNDLRTLEDLLSGIENKTSEAKELIKSIQVSYNPNEPPKPNTIVSATKDSKKMM O02676_CROCR/1-77
RFGSYCPTTCGIADFLSTYQGSIDKDLQTLEDILNQVENKTXEASELIKSIQVSYNPDEPPRPNMIEGATQKSKKML Q6X869_TENEC/1-77
RFGSYCPTTCGIADFLSTYQTKVDKDLQSLEDILHQVENKTSEVKQLIKAIQLTYNPDESSKPNMIDAATLKSRKML FIBG_HUMAN/40-116
RFGSYCPTTCGIADFLSTYQTXVDKDLQVLEDILNQAENKTSEAKELIKAIQVRYKPDEPTKPGGIDSATRESKKML O02689_TAPIN/1-77
RFGSYCPTMCGIAGFLSTYQNTVEKDLQNLEGILHQVENKTSEARELIKAIQISYNPEDLSKPDRIQSATKESKKML O02688_PIG/1-77
RFGSYCPTTCGVADFLSNYQTSVDKDLQNLEGILYQVENKTSEARELVKAIQISYNPDEPSKPNNIESATKNSKRMM O02672_9CETA/1-77
RFGSYCPTTCGIADFLSNYQTSVDKDLQDFEDILHRAENQTSEAEQLIQAIRTSYNPDEPPKTGRIDAATRESKKMM O02682_EQUPR/1-77
RFGSYCPTTCGIADFLSTYQTKVDEDLQNLEDILYRVENRTSEAKELIKAIQVDYNPGEPPKQSVTEGATQNAKKMV Q6X870_CYNVO/1-77
RFGSYCPTTCGISDFLNSYQTDVDTDLQTLENILQRAENRTTEAKELIKAIQVYYNPDQPPKPGMIEGATQKSKKMV FIBG_RAT/40-116
RFGSYCPTTCGIADFLNKYQTTIDQDLRHMEETLRDIDNKTAESTLLIQKIQIGQTPDPRPQ-NVIGDVTQKSRKMI Q6X866_DROAU/1-76
RFGSYCPTTCGIADFFNKYRLTTDGELLEIEGLLQQATNSTGSIEYLIQHIKTIYPSEKQTLPQSIEQLTQKSKKII O93568_CHICK/40-116
RFGEYCPTTCGISDFLNRYQENVDTDLQYLENLLTQISNSTSGTTIIVEHLIDSGKKPATSPQTAIDPMTQKSKTCW FIBG_XENLA/38-114

Alignment Output

As in Bio.SeqIO, there is a single output function Bio.AlignIO.write(). This takes three arguments: some alignments, a file handle to write to, and the format to use.

As of Biopython 1.48, the alignment object acquired a .format() method to give a string containing the alignment in the specified file format, e.g.

AlignIO.read(open("PF09395_seed.sth"), "stockholm")
print(alignment.format("fasta"))

Please refer to the Bio.AlignIO chapter in the Tutorial for more details.

File Format Conversion

Suppose you have a file containing PHYLIP alignment(s) that you want to convert into the PFAM/Stockholm format:

from Bio import AlignIO

input_handle = open("example.phy", "rU")
output_handle = open("example.sth", "w")

alignments = AlignIO.parse(input_handle, "phylip")
AlignIO.write(alignments, output_handle, "stockholm")

output_handle.close()
input_handle.close()

By changing the format strings, that code could be used to convert between any supported file formats.