No available matched normal for Mutect2 GATK

Question

[shems2001]

Hi everyone,

I am attempting the GATK Mutect2 somatic mutation calling pipeline with MC38 CRC WES data, yet I do not have a "matched normal," as the WES data came from a cell line.

Thus, I was advised to use the GRCm38 (mm10) as the reference genome, use my MC38 WES data as the tumor sample, and use the C57BL/6J mouse reference genome as my "matched normal," since I have no other option.

1) Would this be considered a sound approach? Or would I yield better results if I just ran Mutect2 without a matched normal?

2) Also, if the approach is fair, I was also wondering why my bwa-mem2 mem step is taking 8+ hours to run on the C57BL/6J genome, and how to speed it up. This is what I ran:

bwa-mem2 mem -t 4 -R @RG\tID:MATHCED_NORMAL\tPL:ILLUMINA\tLB:ERR9880493" ${reference_genome} ${matched_normal}/ERR9880493_1.fastq.gz ${matched_normal}/ERR9880493_2.fastq.gz > ${aligned_reads}/ERR9880493.paired.sam

My file sizes are approximately 65 GB each. I apologize if this is a rudimentary question, I'm very new to bioinformatics. Any advice would be greatly appreciated.

[Pierre Lindenbaum]

Also, if the approach is fair, I was also wondering why my bwa-mem2 mem step is taking 8+ hours to run on the C57BL/6J genome

split the fastqs in , say, 10 parts, run the mapping in parallel, pipe into samtools sort to produce bam and merge the bams later.

[shems2001]

Solid advice, thank you.

[GenoMax]

My file sizes are approximately 65 GB each

and

bwa-mem2 mem step is taking 8+ hours

Considering the file sizes and the fact that you are using 4 cores, normal.

[Pierre Lindenbaum]

Considering the file sizes and the fact that you are using 4 cores normal.

yes