Quantcast

Rerunning Maker

classic Classic list List threaded Threaded
2 messages Options
| Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Rerunning Maker

Ian Misner
Hello,

I have run maker 2.03 on a genome that we assembled a while back.  Now we have a new genome assembly.  Is there a way to rerun maker with the new assembly without losing all of the annotations I've done on the v2.03 predicted proteins?  I see the genome_gff option but all the contig numbers will have changed with the new assembly so that will not work.  The EST data will be the same as before.  From my original run I used a related species for protein homology, this time would I use the predicted set I have been working with?  Worst case scenario I need a way to link the old proteins with the new ones which I could do with BLAST post Maker but I'm hoping there is a better way.  Any help with this would be greatly appreciated.  


Cheers
Ian

*************************************
Mr. Ian Misner
PhD. Candidate
Department of Biological Sciences
University of Rhode Island
120 Flagg Road
Kingston, RI 02881
Office: CBLS 260
Ph: 1-401-874-9726
fax (401) 874-2065
[hidden email]
http://cels.uri.edu/bio/lanelab/ 








_______________________________________________
maker-devel mailing list
[hidden email]
http://box290.bluehost.com/mailman/listinfo/maker-devel_yandell-lab.org
| Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Rerunning Maker

Carson Holt-2
There is a way to do this and it sounds harder than it really is.

Previously there was a tool that came with MAKER called map2assembly that
is no longer included with the distribution (certain changes to the MAKER
libraries broke it beyond repair).  The replacement for this tool is still
somewhat under development, so is not documented but is included with the
distribution and works far better (and faster) than map2assembly.

The new tool is internal to MAKER and called by setting several flags in
the control files before running a regular MAKER job.

To map genes forward do this:
1. Supply your old transcript file to the est= option in MAKER.
2. Set est2genome=1
3. Set single_exon=1
4. Set single_length=1
5. Turn all repeat masking options off (either delete values in control
files or use -R flag on the command line)
6. Set est_forward=1 (this is not in the maker_opts.ctl file.  You will
have to add it manually.
7. Don't run any gene predictors and don't provide any other evidence
files (we only want MAKER to map things forward)
8. Now run MAKER (remember to supply the -R flag from the command line if
you didn't delete masking options from thew control files)

MAKER will now map the old gene models to the new assembly (with names)
generating a new GFF3 file that can be used as input to future MAKER runs.

Some things to know about the process.  You can set the minimum accepted
coverage and identity to use when mapping the genes forward using the
blastn filters in the maker_bopts.ctl file.  Some genes may not map
forward and will be lost.  Some will map to multiple locations, so review
your GFF3 results file.  Also the est_forward option accepts hints via a
tag (maker_coor=) in the fasta headers.  For example, if you add the
following flag to fasta headers for the input transcript file -->
        >gene1-RA  maker_coor=chr1;
Then MAKER will only try and map that transcript to the new genomes contig
labelled chr1.  Or you can do -->
        >gene1-RA  maker_coor=chr1:2000-200000;
Then MAKER will restrict the alignment to somewhere within a given region
(could be the first half of the chromosome for example).


In future releases of MAKER, everything will be simplified.  The
est_forward will always be in the control files and will accept a file
name.  MAKER will then do all the work upstream including setting flags
and sorting out multiple location alignments as part of it's normal run
(not as an independent run as is done in what I described).

Thanks,
Carson


On 12-03-01 5:25 PM, "Ian Misner" <[hidden email]> wrote:

>Hello,
>
>I have run maker 2.03 on a genome that we assembled a while back.  Now we
>have a new genome assembly.  Is there a way to rerun maker with the new
>assembly without losing all of the annotations I've done on the v2.03
>predicted proteins?  I see the genome_gff option but all the contig
>numbers will have changed with the new assembly so that will not work.
>The EST data will be the same as before.  From my original run I used a
>related species for protein homology, this time would I use the predicted
>set I have been working with?  Worst case scenario I need a way to link
>the old proteins with the new ones which I could do with BLAST post Maker
>but I'm hoping there is a better way.  Any help with this would be
>greatly appreciated.
>
>
>Cheers
>Ian
>
>*************************************
>Mr. Ian Misner
>PhD. Candidate
>Department of Biological Sciences
>University of Rhode Island
>120 Flagg Road
>Kingston, RI 02881
>Office: CBLS 260
>Ph: 1-401-874-9726
>fax (401) 874-2065
>[hidden email]
>http://cels.uri.edu/bio/lanelab/
>
>
>
>
>
>
>
>
>_______________________________________________
>maker-devel mailing list
>[hidden email]
>http://box290.bluehost.com/mailman/listinfo/maker-devel_yandell-lab.org



_______________________________________________
maker-devel mailing list
[hidden email]
http://box290.bluehost.com/mailman/listinfo/maker-devel_yandell-lab.org
Loading...