###PrediXmod ####Pipeline to develop prediction models for GTEx data and cross-validation

-Instructions for downloading expression and genotype data from GTEx exchange area are listed in gtex.md

-Instruction for accessing GitHub from a secure server are given in tarbell_to_github.md

-Descriptions of scripts that execute steps in gene prediction model development pipeline shown below:

Convert GTEx genotypes to downstream file formats

• for all autosomal SNPs combined in one file: 1_vcf2dosage.mach_gtex_hapmapSNPs.pl
• for separate files per autosome: 1_vcf2dosage.mach.chr_gtex_hapmapSNPs.pl

Convert GTEx RNA-seq data for downstream analyses

• 2_pull_protein_coding_RPKM_rebuild.R

Normalize and adjust the GTEx RNA-seq data by PEER factors and genotype principal components

• 3_calc_PEER_adj_exp.r

Run LASSO cross-validation for GTEx tissue of interest, save R² and best betas

• 4_CV_GTEx_lasso_adjusted.r

Run polyscore cross-validation for GTEx tissue of interest, save R² and top betas

• 8_CV_GTEX_polyscore_adjusted.r