Combining normal+tumor fastq files to make a diluted tumor sample

Question

[enho]

Hi everyone,

I have a tumor sample and it's matched normal fastq files. I consider my tumor sample contents to be ~99% pure.

I am trying to dilute this tumor sample to ~10% tumor content. The idea is to make a fastq file with 90% of matched normal reads and 10% of tumor sample reads.

I'd like to keep the read depth as it is, meaning if my pure tumor fastq have a mean depth of 100, after the dilution it still have the same depth.

I was wondering is there any easy way to do this? Would it be easier if I edit the aligned files?

Thanks.

[Kevin Blighe]

You would have to program this in a customised fashion. Python would be the way to go (although, I would find a way to do it via AWK).

[enho]

Hi Kevin, thanks for your reply, I finally managed to do it myself.

What I did was to write a shell script which uses "fastq-sample" tool to uniformly sample for example 90mil reads from normal and then 10mil reads from the tumor. Then write both into one fastq (if pair end, fastq-sample will handle smapling matched reads and then you can write in 2 fastq files) file and then align them.

Thought it might help someone else :)