Feature request: overriding `Reference genome mismatch` error

[evanbiederstedt]

Hello

This issue is related to the following: https://github.com/Illumina/manta/issues/93

I'm configuring Manta as follows (and the bug is the same for germline or TN pairs)

  configManta.py \
    --referenceFasta file.fa \
    --tumorBam tumor.bam \
    --runDir Manta

I'm getting the following error:

Reference genome mismatch: Normal BAM/CRAM file is missing a chromosome found in the reference fasta file: 'NC_007605'

or

Reference genome mismatch: Reference fasta file is missing a chromosome found in the Tumor BAM/CRAM file: 'NC_007605'

So, the input BAMs used a somewhat different reference. Normally, this is an informative error. But for the following analysis, I would like to override this.

Is there a flag to do this? If not, could such a feature exist?

Thanks

[x-chen]

Unfortunately, Manta doesn't support a flag to do chromosome overriding. The QC check was designed to capture reference errors such as hg19 vs. hg38. We may consider a feature to allow chromosome overriding as an enhancement.

[evanbiederstedt]

Hi @x-chen

Thanks for the reply

The QC check was designed to capture reference errors such as hg19 vs. hg38.

Yes, this makes a good deal of sense.

We may consider a feature to allow chromosome overriding as an enhancement.

Excellent, I appreciate this. There are a few bioinformatic analyses whereby I use SV calling against a non-standard reference. I was curious how Manta performed in such a case, even if it is not a "standard procedure". Hacking the source code felt a bit silly, as there are a few reasons why I could envisage this feature being useful.

Thank you for the help, Evan