About promiscuous filter

2 views

Skip to first unread message

小林賢子

unread,

Jun 8, 2024, 5:05:59 AMJun 8

to STAR-Fusion

Dear Dr. Haas

Always thanks.

Today I would like to ask about the 'promiscuous fusion filter'.

You wrote as below.

Promiscuous fusion filter: if candidate GeneA–GeneB exists along with alternative fusion candidates GeneA–GeneC and GeneA–GeneD, and the fusion evidence supporting GeneA–GeneB greatly exceeds that of the alternative fusions (at least 20× support), the alternatives are discarded and the dominant fusion pair is retained. If afterwards, GeneA is found to have at least ten fusion partners, all GeneA-containing fusion pairs are excluded from the sample altogether.

I think this is the way it should be, I guess. However, I don't understand it theoretically.

Are Promiscuous fusions false positives (which are not really present) or are they some kind of algorithmic artifact?

Tetsuko Kobayashi

Brian Haas

unread,

Jun 8, 2024, 8:03:48 AMJun 8

to 小林賢子, STAR-Fusion

Hi,

It could be biological or artifacts - some highly expressed transcripts get trans-spliced into other transcripts, and during cDNA prep and PCR, there's a lot of false priming that can happen among some sequences, resulting in these kinds of findings.

--
You received this message because you are subscribed to the Google Groups "STAR-Fusion" group.
To unsubscribe from this group and stop receiving emails from it, send an email to star-fusion...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/star-fusion/ff493d3c-1e45-4f1a-bc67-128214a810e2n%40googlegroups.com.