STAR-Fusion: Problem in interpreting

[Ranjan J. Sarma]

Could anyone give me an idea about these two terms "JunctionReadCount" and "SpanningFragCount" of STAR-Fusion result?

For some fusion, these numbers are very high (More than 100) and for some fusion, the numbers are extremely low (Less than 20). In this case, is there any threshold for filtering out the fusions having lower JunctionReadCount and SpanningFragCount?

[Brian Haas]

Hi,

The junction reads overlap the fusion breakpoint and provide evidence for the breakpoint site itself.  The spanning fragment reads don't overlap the breakpoint position, but have read pairs on either side of the breakpoint (spanning the breakpoint).   The number of junction vs. spanning reads usually depends on both the length of the reads (longer the reads, more likely a junction read type) and position of the breakpoint in the context of the full length fusion transcript (if the breakpoint is localized towards either end of the fusion transcript, more likely to be a junction read).

Modern versions of STAR-Fusion automatically filter based on the total number of supporting reads, requiring at least 1 fusion read per 10 M total reads (0.1 fusion fragments per million (ffpm)).

best,

~b

On Mon, Aug 19, 2019 at 1:39 AM Ranjan J. Sarma <ranjan...@gmail.com> wrote:

Could anyone give me an idea about these two terms "JunctionReadCount" and "SpanningFragCount" of STAR-Fusion result?

For some fusion, these numbers are very high (More than 100) and for some fusion, the numbers are extremely low (Less than 20). In this case, is there any threshold for filtering out the fusions having lower JunctionReadCount and SpanningFragCount?

--
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/755dcd9b-bd09-421c-b82f-2d10ea183d9f%40googlegroups.com.

--

--
Brian J. Haas
The Broad Institute
http://broadinstitute.org/~bhaas

 

[Ranjan J. Sarma]

This is helpful.  Thank you so much, Sir.

...
Regards,

Ranjan Jyoti Sarma
Junior Research Fellow
Department of Biotechnology,
MZU, Aizawl, India

[Dicer]

Having read the descriptions of `JunctionReads` and `SpanningFragCount` in docs, I assume that JunctionReads is the preferred quantification since the fusion is actually observed.

[Brian Haas]

The junction reads give you the breakpoint and so provide the highest
specificity. The longer the reads, they eventually all turn into
junction reads, too.

> To view this discussion on the web visit https://groups.google.com/d/msgid/star-fusion/2b18bd8e-c2b8-4f8c-8980-31d63591dfaen%40googlegroups.com.