Workflow for eQTL analyses in the developing human brain

• This workflow consists of a series of Snakemake workflows that can be executed using the supplied bash scripts:

• Genotyping.sh: run initial genotyping QC and prepare data for imputation

• eQTL.sh: QC and filtering of post-imputation genotypes and expression data; eQTL analysis

• GTcheck.sh: map reads to reference, call snps, and compare to imputed genotypes

  • This needs to be run after eQTL analysis because it requires processed vcf files from eQTL workflow

• SMR.sh: SMR analyses of eQTL data (requires config file that will be made available on publication)

• Expression data were created from the workflows detailed here

• Software used:

  • bcftools v1.6
  • bedtools v2.26.0
  • bgzip v1.3.1
  • checkVCF.py v1.4
  • CrossMap v0.2.5
  • fastQTL v2.184_gtex
  • hisat2 v2.0.5
  • pandas v0.21.0
  • PEER v1.3
  • PLINK v1.90b3.39
  • samtools v1.3.1
  • smr v0.69
  • tabix v1.3.1
  • tidyverse v1.1.1
  • vcf-sort

• Genome References:

  • variant lookup table in VCF format (dbSNP v149, genome build GRCh37.p13)
  • hg19 to GRCh38 chain file
  • genome sequence (GRCh38.p4 with Decoy sequences)
  • splicesites (derived from Gencode version 23 (GRCh38.p3))
  • 1000 Genomes Project Phase 3 genotypes:
    • PLINK files
    • vcf files

• GTEx data:

  • SNP lookup table (GTEx_Analysis_2016-01-15_v7_WholeGenomeSeq_635Ind_PASS_AB02_GQ20_HETX_MISS15_PLINKQC.lookup_table.txt.gz)
  • significant eQTLs (GTEx_Analysis_v7_eQTL.tar.gz)
  • Metasoft multi-tissue eQTL data (GTEx_Analysis_v7.metasoft.txt.gz)