SNAP produced a long long long gene(~40M)?

[xiekunwhy]

Hi,

SNAP ofthen give some strange results, for example, output a long long long gene, the length of the gene cover almost the whole scaffold, why and how to avoid that?

Here is the hmm file I used and the result file SNAP output.

snap.zip

Best, Kun

[iankorf]

There was a problem in training. Your HMM can't make initial exons.

[xiekunwhy]

Hi,

I used following commands to run training and it seems that every things worked OK, I don't known why hmm file has problem,

perl gff3_to_zff.pl -g mikado.keep3.gff -f Op-f.pctg.fa -o mikado.keep3.zff fathom mikado.keep3.zff Op-f.pctg.fa -validate > validate.log perl cleanZff.pl -z mikado.keep3.zff -l validate.log -o mikado.keep3.clean.zff fathom mikado.keep3.clean.zff Op-f.pctg.fa -categorize 1000 > categorize.log fathom uni.ann uni.dna -export 1000 -plus > uni-plus.log mkdir params cd params forge ../export.ann ../export.dna > ../forge.log perl hmm-assembler.pl Op-f params > Op-f.hmm

Best

[xiekunwhy]

Maye the gff3_to_zff.pl script gives a wrong guide, it only produce Exon region.

[iankorf]

It might be a good idea of we did a zoom. Send me your availability.

From: @.> Sent: Sunday, December 26, 2021 6:37 PM To: @.> Cc: Ian @.>; @.> Subject: Re: [KorfLab/SNAP] SNAP produced a long long long gene(~40M)? (Issue #13)

Maye the gff3_to_zff.pl script have a wrong guide, it only produce Exon region.

— Reply to this email directly, view it on GitHubhttps://github.com/KorfLab/SNAP/issues/13#issuecomment-1001299170, or unsubscribehttps://github.com/notifications/unsubscribe-auth/AAI6KN4FO4OGJK5X5PWOGQLUS7GQJANCNFSM5KYXHVMA. You are receiving this because you commented.Message ID: @.***>

[xiekunwhy]

Here is my script, but I need to read all gtf and fasta file to RAM. gtf2zff.zip

[xiekunwhy]

And here is snap zff to gtf script，those who want gff can use gffread to convert gtf to gff snap2gtf.zip