How to use PASA to add novel gene from RNA-seq alignment to existing gene feature files ( GFF )?

Question

[Maimai]

Hi,

I'm new for the PASA software. I've been dealing with gene annotation in drosophila recently and I faced a problem when dealing transcriptome annotation with PASA. The question is about the TRANSDECODER parameter. When I added this parameter in the step "Performing an annotation comparison and generating an updated gene set" of the script ../scripts/Launch_PASA_pipeline.pl, it didn't add novel gene comparing to the one without adding TRANSDECODER parameter. The thing is, I am wondering if it can help me to find new gene of from transcript alignment that the existing annotation don't have and add those genes to updated GFF files.

Best, Maimai

Answer 1

[Juke34]

Yes I confirm you it can help you to find new genes from transcript alignments as well as new isoforms for gene already annotated. Could you share your command line ?

[Maimai]

Here's the command line: nohup $tool/PASApipeline-2.0.2/scripts/Launch_PASA_pipeline.pl -c annotCompare.config -A -g genome.fasta -t existing_annotation.gff3 --TRANSDECODER &

Thank you so much for ur reply~~

[Juke34]

You have to provide the annotation you want to update with the option --annots_gff3 (i.e the help page). You used instead the -t that is to provide the transcripts file to use to perform the new predictions (in fasta format). So obviously is missing this latest one.

[Maimai]

I changed my command: nohup $tool/PASApipeline-2.0.2/scripts/Launch_PASA_pipeline.pl -c annotCompare.config -A -L --annots_gff3 existing.annotation.gff3 -g genome.fasta -t transcript.fasta --TRANSDECODER& But the output still have no novel gene been added, there are some revised on adding UTR though.

The option --annots_gff3, should it be added in the step of transcript alignment or annotation comparison ?

Thank you so much !

[Juke34]

I guess it's because you have to use on top of the parameters already used, the "-R" and "--ALIGNERS blat,map" to launch the alignment pipeline to pre-process your transcript.fasta. Otherwise you transcript.fasta file is useless...

[ashaneev07]

Hii... I got the following error while running PASA . Could u please help me out of this?

./Launch_PASA_pipeline.pl -c conf.txt -C -R -g test.pasa.fa -t Unigenes.fasta --ALIGNERS blat,gmap
    Error, couldn't extract DATABASE from config file /EVM/PASA/PASApipeline.v2.4.1/conf.txt