[maker-devel] Extract FASTA Sequences from "Maker Standard" Build

[Michael Campbell]

Hi Jason,

Rerunning MAKER with the standard gff3 file would work, but for speed I would use the fasta_tool accessory script that is bundled with MAKER. All you need to make is a file with the list of transcript names from the standard gff3. Then you can use fasta_tool with the --select ooption to return all of the FASTA sequences that are in the list. The command would look like this

PATH_TO_MAKER/maker/bin/fasta_tool --select id_file.txt max_transcritps.fasta | PATH_TO_MAKER/maker/bin/fasta_tool --wrap 80 > standard_transcripts.fasta

fasta_tool outputs unwraped fasta by default, so I generally pipe the output back through fasta_tool to wrap the text. The above command line wraps the sequence at 80 characters.

you can use a perl one liner like this one to make the id file

perl -lane ' if ($F[2] eq mRNA){my ($id) = $_ =~ /Name=(\S+?);/; print $id;}’ maker_standard.gff

If you use these command line make sure you type them out yourself, email programs have a tendency to change characters slightly making copy/pasted command fail.

Thanks,
Mike  

> On Feb 9, 2016, at 9:36 PM, Jason Gallant <jgallant at msu.edu> wrote:
> 
> Hi Everyone,
> 
> Quick question— I’ve run through Mike Cambell’s tutorial on building “Maker Standard”, “Maker Default” and “Maker Max” datasets.  I’ve decided that the “Maker Standard” data (Transcripts with Evidence and/or IPR scan hits) makes the most sense for what we’re trying to do.
> 
> Is there an easy way to create the fasta files associated with the maker standard build?  Fasta_merge typically outputs a variety of .fasta files, which I’ve been able to create following this protocol for the “maker max” dataset.  I’d like to get these for the “maker standard” build.
> 
> Currently, the datastore contains the data for the “maker max” data.  One way, i suppose would be to re-run MAKER with the maker standard gff file, but it seems like an overly complicated way of doing it…?
> 
> Any suggestions Mike (or others?)  Has anyone written a script to do this automagically?
> 
> Best,
> Jason Gallant
> _______________________________________________
> maker-devel mailing list
> maker-devel at box290.bluehost.com
> http://box290.bluehost.com/mailman/listinfo/maker-devel_yandell-lab.org