[maker-devel] calculating AED values between two datasets

[Carson Holt]

If you got every value as 1 with the est_gff, then your GFF3 didn’t load.  The est_gff option is expecting match/match_part format alignment format, and you may not have had it correctly structures.

For using fasta files instead, you may also need to set single_exon=1 and single_length=1, otherwise many of those alignments will be ignored for AED scoring purposes.  You should also look out the output in a viewer like apollo to visualize the comparison to see if the reason you get 1 is because the aligner can’t recover the original transcript alignment.

—Carson



> On Nov 7, 2014, at 6:17 AM, Poelchau, Monica <monica.poelchau at ars.usda.gov> wrote:
> 
> Hi everyone,
> 
> I would like to generate a list of Maker AED values comparing two datasets: a set of computationally predicted genes, and manually curated genes from the Web Apollo program. The idea is to quantify the amount of nucleotide-level change that occurred during the manual curation process. 
> 
> I have tried to run Maker in several ways to generate the AED values. Both gene sets are in (as far as I can tell) valid gff3 format. First, I included the manually curated (Web Apollo) gff3 in the 'model_gff' field of maker_opts.exe, and the gff3 of the computational predictions in the 'est_gff' field, with all of the other prediction and evidence alignment settings turned off. All resulting AEDs from this analysis were 1, even though many of the annotations had 100% overlap. Next, instead of using the computational predictions in gff3 format, I used the fasta file of the cDNA sequence from the computational predictions in the 'est' field. Here, the results made more sense, but there was a small but significant percentage of the AED values that were 1 that actually should have been less than 1. 
> 
> I have tried the 2 analyses above using both the gff3 output straight from Web Apollo, and after running the gff3 through maker once as the only entry in the model-gff field, as explained in the MAKER2 paper (http://www.biomedcentral.com/1471-2105/12/491). This does not to appear to make a difference.  
> 
> Do you have any ideas where I might start to debug this?
> 
> Thanks for your help!
> 
> Monica
> 
> 
> 
> 
> 
> This electronic message contains information generated by the USDA solely for the intended recipients. Any unauthorized interception of this message or the use or disclosure of the information it contains may violate the law and subject the violator to civil or criminal penalties. If you believe you have received this message in error, please notify the sender and delete the email immediately.
> _______________________________________________
> maker-devel mailing list
> maker-devel at box290.bluehost.com
> http://box290.bluehost.com/mailman/listinfo/maker-devel_yandell-lab.org

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://yandell-lab.org/pipermail/maker-devel_yandell-lab.org/attachments/20141107/92afdf9e/attachment-0003.html>