How can identify the driver genes with germline mutations ?

Question

[andy7v9591]

Hi all,

I am working on DNA data (our own experimental data) for some cancer types of human. We only have Fastq data from normal sample without pair tumor sample. I have done variant calling with GATK Tool and get the vcf file (Germline mutations). However, most of cancer research compare information (somatic mutations) about cancer and germline DNA. I want to identify the driver genes (patients can vary the type and severity of possible side-effects) with germline mutations. Now, I only can do some routine analysis (i.e. calculate mutation frequency, annotate variants to genomic features, assess the functional effect, ...). About some tools be used to identify driver genes from a cohort of cancer patients such as MutSigCV, MuSiC, OncodriveFM. Do these tools make the right? Or what is the right tool for the analysis germline mutations? Could anyone share some experience about this issue?

Many thanks !

Andy - Lin

Supplement information

Pharmacogenomics (a portmanteau of pharmacology and genomics) is the study of the role of genetics in drug response. Most patient populations show large inter-individual variability in drug response and toxicity (i.e., certain drugs do not work as well as expected, whereas in other patients they cause toxic effects, even at lower doses). For some patients, the reason may be genetic.

I have two group of patients. The first group doesn't develop toxic effects. The second group develops toxic effects.

I want to identify the driver genes causing toxic effects (not causing cancer) with germline mutations (SNP, InDel).

[Sean Davis]

Could you clarify your biological question? Are you looking for risk alleles for developing cancer? Or are you looking for potential drug response modifiers? Something else?

[andy7v9591]

Sorry about my unclear description. I do some supplementary.

[Sean Davis]

So, how many samples in each group? Are these exomes or genomes? Do you know the drug that likely caused the toxicity?

[andy7v9591]

There are only 10 sample per group.

Data is WGS sequencing data.

Because not everyone who undergoes treatment develops adverse side effect, I want to discovery genetic predisposition to human disease.

Answer 1

[Jimbou]

If your samplesize is high enough you can apply a genomewide case control study using tools like plink. For small samplesize you can find out the pharmacokinetik pathway of the given drug and search for functional variants in the involved genes.

[andy7v9591]

Thanks for your replying.

Our sample size is small because of high cost of WGS. Your advise is a good point. I can screen a group of candidate gene from pharmacological network.