Best Approach for Variant Calling with Mutect2 on Drug-Sensitive and Drug-Resistant Cancer Cell Lines

[jenny8030]

First of all, thank you for developing and maintaining this tool! It has been incredibly helpful for my research.

I plan to perform variant calling using Mutect2 and would like your advice on the best approach. Below are the details of my samples and analysis goal:
- My samples are derived from the human cancer cell line A, with one being drug-sensitive (A-S) and the other drug-resistant (A-R).
- I aim to identify the differences between these two samples.

For the analysis, I’m considering two approaches:
1. Running Mutect2 in tumor-only mode separately for A-S and A-R, and then comparing the results to identify the differences.
2. Treating A-S as the normal sample and A-R as the tumor sample, and running Mutect2 in paired mode.

Could you provide your opinion on which approach would be more appropriate for this analysis? Any additional recommendations are also appreciated. Thank you!

[gokalpcelik]

Hi @jenny8030
The latter approach is better since Mutect2 is able to use any sample as pseudonormal therefore our recommendation would be to do it in paired mode. You may be able to expand your options if you can also try the paired mode in reverse and see if there are differences that may cause the differences between 2 cell lines.

Regards.

[jenny8030]

Thank you for your recommendation! I will proceed with the paired mode as suggested.

I do have a follow-up question:

• A-S is not truly a "normal" sample but rather a tumor cell line. This makes me a bit concerned.
• Does Mutect2 assume that the normal sample is diploid or have other specific assumptions about the "normal" sample?

I want to make sure that using A-S as the "normal" sample won’t introduce any unexpected biases in the analysis.

Your guidance is greatly appreciated!

Thank you!

Best regards.

[gokalpcelik]

Mutect2 has no assumptions about normal or tumor samples therefore using any cell line as pseudonormal should not result in any unexpected outcomes.