01a-perSiteGLMM.R: For each SNP (introgressed or nonintrogressed), fit a generalized linear mixed model to estimate the allelic effect. Then determine significance at 10% FDR.
01b-perSiteGLMM_tissue_heterogeneity.R: For each SNP, fit a GLMM with and without a fixed effect of tissue. Calculate the Bayes factor to compare the two models.
02-getDerivedAF.sh: Extract the ancestral/derived allele calls from the 1000 genomes dataset.
03-sigByDAF.R: Compare the proportion of introgressed and nonintrogressed SNPs showing significant ASE, stratifying by derived allele frequency to control for power. [Fig. 2, Fig. 4a, Fig. 4b]
04a-aseByTissue.R: Fit a GLMM to the full introgressed dataset (rather than per-SNP) with tissue as a fixed effect. Compare the coefficient estimates for different tissues. [Fig. 5a]
04b-aseByTissue_matched_control.R: Fit the same model to equal-sized samples of covariate-matched non-introgressed control SNPs to further evaluate signficance of downregulation of Neanderthal alleles.
05_sigByTissue.R: Compare proportions of up- and down-regulated SNPs per tissue. [Fig. 5b]
06_divergenceByTissue.R: Get the expression-weighted divergence between modern human and Neanderthal gene sequences per tissue. [Fig. 5c]
GTEx Portal http://www.gtexportal.org/
GTEx dbGaP Page
http://www.ncbi.nlm.nih.gov/projects/gap/cgi-bin/study.cgi?study_id=phs000424.v6.p1
Altai Neanderthal VCFs
http://cdna.eva.mpg.de/neandertal/altai/AltaiNeandertal/VCF/
1000 Genomes VCFs
ftp://ftp.1000genomes.ebi.ac.uk/vol1/ftp/release/20130502/