Is there a hard limit on inter-exonic ranges?

[markb729]

Does collapse_isoforms_by_sam.py have an upper limit to the inter-exon distance it will collapse into a putative isoform? In my animal model, I have a wildtype allele that spans about 154 Kb (some 40 or so exons) and generates a 8.6 Kb transcript. I also have engineered a deletion that spans approximately 83 Kb in a heterozygote background, which yields a much smaller transcript comprised of about 12 exons. This is easily seen with FLNC transcripts splice-mapped to the reference, shown below, including a majority of reads exhibiting the deletion (NB: the coverage imbalance is due to a stronger promoter on the deletion and not all reads with the deletion are shown).

Consisting about 70 percent of all FLNC reads, this deletion should have been easily detected. However, the output of **collapse_isoforms_by_sam.py --dun-merge-5-shorter ... ** yields only three slightly different patterns of the wild type allele; the KO allele is completely absent. No filtering or other 5' contingencies were performed. Do you have any explanation of why this may have occurred?

Thanks.

[Magdoll]

Hi, there is no inter exon distance limit. My guess is those KO isoforms but filtered out by collapse due to either alignment coverage below cutoff (99%) or identity (95%). When you run collapse you get a “ignored-ids.txt” which tells you which FLnC reads were excluded and why — since you already know the FLnC for those KO can you see what happened to them?

On Thu, Feb 10, 2022 at 11:33 AM markb729 ***@***.***> wrote: Does *collapse_isoforms_by_sam.py* have an upper limit to the inter-exon distance it will collapse into a putative isoform? In my animal model, I have a wildtype allele that spans about 154 Kb (some 40 or so exons) and generates a 8.6 Kb transcript. I also have engineered a deletion that spans approximately 83 Kb in a heterozygote background, which yields a much smaller transcript comprised of about 12 exons. This is easily seen with FLNC transcripts splice-mapped to the reference, shown below, including a majority of reads exhibiting the deletion (NB: the coverage imbalance is due to a stronger promoter on the deletion and not all reads with the deletion are shown). [image: Screen Shot 2022-02-10 at 12 53 24 PM] <https://user-images.githubusercontent.com/15279264/153478210-85029e1f-0463-442a-af94-75cafe7ce354.png> Consisting about 70 percent of all FLNC reads, this deletion should have been easily detected. However, the output of **collapse_isoforms_by_sam.py --dun-merge-5-shorter ... ** yields only three slightly different patterns of the wild type allele; the KO allele is completely absent. No filtering or other 5' contingencies were performed. Do you have any explanation of why this may have occurred? Thanks. — Reply to this email directly, view it on GitHub <#194>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAEQE34A6DCA6MBM5CNQU63U2QHH3ANCNFSM5OBZO6PQ> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>. You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

-- Sent from Gmail Mobile. Excuse any possible typos.