I have a RNAseq + human WGS data and I want to do sQTL analysis using QTLtools. My group design includes positive group ( which has 2 severity sub-phenotypes) and negative group.

I have created the following files for use in QTLtools for performing cis permutation pass:

1. phenotype file : intron-usage ratio from leafcutter
2. genotype file : vcf file from WGS data
3. covariate file : PCs from phenotype data and genotype data and age, gender and batch classification.

Q) My question is for capturing severity-specific sQTL what type of covariate file should I have?

For example, I have run it separately for positive group and negative group and captured some sQTLs which I annotated and overlapped to see if they are separate or similar in both the groups.

Q) To get severity-specific QTL should I include that in covariate file for positive group or run QTLtools separately for both the severe group and mild group?

I need help with understanding if my grouping approach for QTL analysis is correct or not, as its my first time doing it.

Thanks