For GTEx V7 data we've done imputation and rna-seq normalization, and we have PCA analysis results from Broad. PEER analysis is still running. Once we have the result we should have the material to do single tissue BVS via mr-ash.
Plan for next steps are:
Work with @pcarbo and @NKweiwang to verify via simplifed toy data that varbvs implementation of mr-ash agrees with mvash. Particularly we need to make sure we have all the quanities we need. Will open a separate ticket for that.
Put together a varbvs based GTEx eQTL pipeline. Inputs are genotype, rna-seq and covariates plus cis-gene definitions. Will dump everything from varbvs for now and decide what to extract and how to make plots, later.
At the same time:
Create simulated data using David's RNA-seq simulator and V7 genotypes.
Also run univariate analysis: via FastEQTL -- reason it might be a bad idea to wait for Broad's official release on this is that result might be slightly different because of different data preprocessing.
For GTEx V7 data we've done imputation and rna-seq normalization, and we have PCA analysis results from Broad. PEER analysis is still running. Once we have the result we should have the material to do single tissue BVS via mr-ash.
Plan for next steps are:
varbvs
implementation ofmr-ash
agrees withmvash
. Particularly we need to make sure we have all the quanities we need. Will open a separate ticket for that.varbvs
based GTEx eQTL pipeline. Inputs are genotype, rna-seq and covariates plus cis-gene definitions. Will dump everything fromvarbvs
for now and decide what to extract and how to make plots, later.