Hi all
I am trying to do lncRNA prediction using toxoplasma gondii RNAseq data. there is this one NAT lncRNA of my interest on anti-sense (NAT lncRNA). Whenever i am using
FEELnc_filter.pl -i assembled/merged_all18_strand.gtf -a reference/ToxoDB-68_TgondiiME49.gtf -s 200 --**monoex=-1** -p 8 -o feelnc_out/filter.log > feelnc_out/candidate_lncRNA_filter.gtf ,
i am not getting that lnc its is getting filtered out. But if i use monoex=1 i get my lnc of interest on antisense.
FEELnc_filter.pl -i assembled/merged_all18_strand.gtf -a reference/ToxoDB-68_TgondiiME49.gtf -s 200 --**monoex=1** -p 8 -o feelnc_out/filter.log > feelnc_out/candidate_lncRNA_filter.gtf
I am not sure why this is happening. if anyone can help me out it will be really nice. Thank you
P.S. - CPC2 is labelling it as coding with score Transcript length 2496 bp
ORF amino acids 149 aa = 447 bp
Fickett score 0.354
ORF completeness 1 (has start+stop) Driving coding call
Final score 0.624 Just above 0.5 threshold
•
written
by
Anurag •
1 answer
I'm not familiar with FEELnc, but it seems to me that the difference between the top and the bottom command is probably that in the top command you are filtering out mono-exonic lncRNAs, where as in the bottom you are filitering out mono-exonic lncRNAs.
In general, we are more confidence that lncRNAs with multiple exons are more likely to be genuine lncRNAs, and not just transcriptional noise.
That is not to say that there are not genuine lncRNAs with only one exon, but just that we can be more confident when splicing is present, where as a lot of non-spliced transcription is just noise.
Of course all of this is tailored for mammalian genomes, where the majority of transcripts have multiple splicing events, and it might not be appropriate for organisms with less splicing, or transcript types that are less likely to be spliced (such as NATs).
Log in to answer this question.