How to approach QTL for transcriptomic data?

Question

[niccolo.fran]

My bioinfromatic group is starting to get into QTLs, unfortunately the only guy with limited experience into the subject left a shrot while ago and hasn't been able to help us.

We have phenotipc data from about 50 individuals each with transcriptomics NGS data and about 120000 SNPs on 10000 genes. I could really use some suggestion on how to approach this, one colleague even questioned if the limited number of individuals would make the analysis pointless as any results would not be significant...

[GouthamAtla]

50 individuals with transcriptome would be enough to get decent number of eQTLs. Which tissue are you working with ? Ideally you should genotye the DNA from these 50 individuals and then do imputation to get more SNPs and then run them through eQTL pipelines-

[niccolo.fran]

The tissue is blood. What kind of pipelines would you recommend?

[GouthamAtla]

It depends. You could simply start with QTLtools or MatrixQTL which are easy fast to run. You could also go to more robust pipelines like RASQUAL which also models for allelic counts.

Answer 1

[EagleEye]

To get more insight into eQTL approaches, these articles will be helpful to you.

1) Genetic effects on gene expression across human tissues

2) Using an atlas of gene regulation across 44 human tissues to inform complex disease- and trait-associated variation