Step 6. Mapping with TensorQTL - ariadnacilleros/Cis-mQTL-mapping-protocol-for-methylome GitHub Wiki

Step 6.1. Compute mQTLs with the main model

In this protocol to map cis-mQTLs, we use TensorQTL, which is based on the statistics from FastQTL but operates in a much faster way and works from a python3 module. It basically performs linear regressions between genotypes and methylome (with or without covariates) in order to find all the nominal associations for each CpG. As the maximal distance spanned by CpG-SNP pairs to be considered for testing, we use the default of 0.5Mb up and down-stream. And, as previously said, we include the sex, the principal components and Planet's cell proportions as covariates in the analysis.

Step 6.2. Compute mQTLs with RNT values

As we have previously mentioned, we will perform an additional model where the methylation values will be the RNT values, and the covariate file will include the sex, the genotype PCs, the Planet cell-type estimations, and the residualized mPCs. All the other parameters will be the same as the main model; ±0.5Mb as the maximum distance spanned by CpG-SNP pair. Attention, this model will be called RNT in the prefix variable from the TensorQTL code, for more details have a look at the corresponding step from the ReadMe.