No CNV remains after filtering

[petrsh]

Follow-up to #57

I’m running a pilot experiment with the objective to select the best tool for distinguishing between normal and malignant cells based on their CNV profile. I use data from this study 10.1016/j.molcel.2021.10.013.

I run sample SRR13914002 with hca_ref and with a custom ref using normal cells from the dataset. In both cases I got this:
"No CNV remains after filtering - terminating."

(I got same comment for another sample from so far six tested samples)

Here are log files -
log_custom_ref.txt
log_hca_ref.txt

Bulk plots for custom ref:

Bulk plots for hca ref

This is result from Copykat:

where ‘not available’ are malignant cells classified based on particular marker genes.

[teng-gao]

Thanks for this report. There seems to be a problem in the pileup_and_phase stage - many SNPs appear to be homozygous and somehow included in the allele profile. Is there a way you could share with me the BAM and barcode list so I can reproduce this?

[petrsh]

Thanks for quick replies! Unfortunately I'm afraid that I can't as access to these raw data is restricted...I have to check with my supervisor first.

[teng-gao]

I see. I think others have reported a similar problem before .. I'm happy to connect via a quick call today or tomorrow to know more about how the genotyping was performed. Let me know; My email is tgaoteng@gmail.com

[teng-gao]

Hi @petrsh,

I suspect your 10x bam was multiplexed and contains multiple individuals - please see this thread here.

[petrsh]

I ran it incorrectly. I set as a label a study ID and put all samples IDs and bams into one run of pileup_and_phase.R...however after running it correctly a problem with 'No CNV remains after filtering' remained. Setting max_entropy = 1 solved that. Thank you for your assistance!