Clustered Events filter is still not smart enough not to filter real somatic events.

[gokalpcelik]

Hi Team (@davidbenjamin )

I've just encountered a strange filtering event due to clustered events filter in one of my samples (though lab says there may be more).

A single INDEL within PTEN gets filtered with Clustered Events although the other events within those reads were already filtered due to being in PON, multiallelic and normal artifact.

Expected output should not filter that PTEN mutation that is only found in Tumor sample.

I am planning on increasing the number of max events in haplotype and region parameters but I am afraid that may inflate the number of FPs.
The older MUTECT2 is around version 4.3 though I used the same version of FilterMutectCalls for both call sets (4.6.2.0). Looks like ECNT count is affecting the way PTEN INDEL gets filtered.

Thank you for the help.