Contamination Estimation Option for Samples with cross-contamination

[maotian06]

The current Mutect2 pipeline can accept an extra option in the input.yaml, here is an example:

  tumor:
    - id: 'TWGSAMIN000001-T001-S01-F'
      BAM: path/to/bam
      contamination_table: path/to/calculatecontamination.table

This will filter some sSNV and potentially increase the accuracy of the Mutect2 caller when you have a sample with high cross-contamination. Here is the table shows the comparison result:

sample	contamination_level	filtered sSNV with contamination-estimation-table	Filtered Out Variant Counts
MTOSIMHO000033	0.13	28.8%	177
MTOSIMHO000034	0.01	7.7%	28
MTOSIMHO000035	0.58	9.1%	842

@tyamaguchi-ucla The reason why the proportion doesn't match the contamination level is that the total variants from different samples are different.
MTOSIMHO000035 has much more variants than MTOSIMHO000033 and MTOSIMHO000034. I added the Filtered Out Variant Counts column here. I can look into the raw code to see how Mutect2 process the contamination level later.