Convert Clustal To Nexus Format In Python

Question

[Gergana Vandova]

I would like to convert a 'clustal' formatted alignment into a 'nexus' format alignment in python. I then would like to use the nexus file to build a phylogenetic tree using Mrbayes (http://mrbayes.sourceforge.net/). How can I do that?

[Neilfws]

Does it have to be Python? There are lots of conversion utilities already available, both online e.g. here and command-line e.g. here

[Gergana Vandova]

I would like to do this conversion for many files, so I wanted to automate it with a script. As I am using python for the multiple sequence alignments, it would be nice if I can use python for the format conversion as well. The Nexus created by NCLconverter that you also recommended is not usable by MrBayes 3.1.

[Adrian Pelin]

trimal does an excellent job at file conversions, it is command line and can be incorporated into your script. It is designed for trimming alignments, but if you omit trimming options, it merely can be used as a convertor.

[Gergana Vandova]

@akoik063 What would be the command for just the conversion? I am checking it out.

[Adrian Pelin]

trimal -in $file -out $file <options>

And part of options is:

   -clustal                 Output file in CLUSTAL format
    -fasta                   Output file in FASTA format
    -nbrf                    Output file in NBRF/PIR format
    -nexus                   Output file in NEXUS format
    -mega                    Output file in MEGA format
    -phylip3.2               Output file in PHYLIP3.2 format
    -phylip                  Output file in PHYLIP/PHYLIP4 format

So, from clustal to nexus: trimal -in $file -out $file -nexus

[Gergana Vandova]

Trimal works! Thanks!

[Neilfws]

Should be easy enough to incorporate an existing command-line tool into your script; seqret from the EMBOSS suite may be another option. I didn't recommend NCLconverter, it was just an example.

[Gergana Vandova]

Do you recommend seqret?

[Neilfws]

I recommend the EMBOSS suite in general, as a great set of small tools, easily incorporated into scripts and pipelines in the "UNIX tradition". I know seqret is a good converter; as to whether it does exactly what you want, you'll have to try it and see!

Answer 1

[gammyknee]

With Biopython you can (assuming you have it downloaded and installed)

$ python
> from Bio import AlignIO
> AlignIO.convert("file.clustal", "clustal", "file.nexus", "nexus")

[Gergana Vandova]

It doesn't work: ValueError: Need a DNA, RNA or Protein alphabet

[David W]

try adding from Bio import Alphabet then adding the argument alphabet=Alphabet.generic_dna to AlignIO.convert

Answer 2

[janwillem.wijnands]

It is possible to delete the quotes in the headers (Seq_name) of the output nexus file?
I would like to use AlignIO from BioPython for this approach.

Example:

#NEXUS
begin data;
   dimensions ntax=3 nchar=50;
   format datatype=dna missing=? gap=-;
matrix
'Seq_name'           AGGGA
'Seq_name'           AGCGG
'Seq_name'           ACTGG

MrBayes gives me an error because of the quotes.