Local adaptation and archaic introgression shape global diversity at human structural variant loci

Supporting data associated with the manuscript "Local adaptation and archaic introgression shape global diversity at human structural variant loci". These include:

• structural variant genotypes (Paragraph; https://github.com/Illumina/paragraph)
• eQTL mapping results (fastqtl permutation pass; see http://fastqtl.sourceforge.net/ for column descriptions)
• eQTL fine-mapping results (CAVIAR; see http://genetics.cs.ucla.edu/caviar/index.html)
• structural variant selection scan results (Ohana; https://github.com/jade-cheng/ohana)

Description of files in this directory:

Structural variant genotypes

SVs_paragraphFormat.vcf.gz - merged long-read structural variant calls

SVs_1KGP_pgGTs.vcf.gz - genotypes for 1000 Genomes samples in VCF format

eQTL mapping results

fastqtl_out.txt - results from fastQTL permutation pass; see http://fastqtl.sourceforge.net/ for column descriptions

caviar_out.txt - results from fine-mapping SNPs and SVs at significant SV eQTL loci with CAVIAR. Description of columns:

• query_sv: SV that was a significant eQTL and underwent fine-mapping
• gene_id: gene exhibiting an expression association with the query_sv
• var_id: variant (SNV or SV) that was tested for expression association with the above gene in the fine-mapping analysis
• var_in_credible_causal_set: Boolean variable denoting whether the above variant is in the 95% credible causal set
• prob_in_pcausal_set: the amount that this variant contributes to 95% credible causal set
• causal_post_prob: the posterior probability that the variant is causal in the expression association

Structural variant selection scan results

chr21_pruned_50_Q.matrix - admixture proportion matrix (generated by Ohana; https://github.com/jade-cheng/ohana)

chr21_pruned_50_F.matrix - matrix of inferred ancestral allele frequencies (generated by Ohana)

chr21_pruned_50_C.matrix - matrix of ancestry component covariances (generated by Ohana) Entries of the matrix can be modified to produce "selection hypothesis" matrices where allele frequencies are allowed to vary in one ancestry component (https://github.com/jade-cheng/ohana/wiki/Population-or-ancestry-specific-selection-scan).

selscan_50_k8_p*.txt.gz - raw output of Ohana selscan (see https://github.com/jade-cheng/ohana)

selscan_res.txt.gz - Ohana selection scan results. These results have been filtered to exclude SVs that have low genotyping rates (<50% of samples), violate Hardy-Weinberg equilibrium expectations (excess of heterozygotes) in more than half of populations, or have extreme global log likelihood estimate (LLE) values. Description of columns:

• ID: SV ID
• #CHROM: SV chromosome
• POS: SV start position
• SVLEN: SV length (negative for deletions)
• step: number of steps needed to interpolate between genome-wide and selection hypothesis models
• lle_ratio: likelihood ratio statistic (LRS) of the genome-wide vs. selection hypothesis model
• global-lle: log likelihood of the genome-wide model
• local-lle: log likelihood of the selection hypothesis model
• f-pop0: inferred allele frequency in ancestry component 0
• f-pop1: inferred allele frequency in ancestry component 1
• f-pop2: inferred allele frequency in ancestry component 2
• f-pop3: inferred allele frequency in ancestry component 3
• f-pop4: inferred allele frequency in ancestry component 4
• f-pop5: inferred allele frequency in ancestry component 5
• f-pop6: inferred allele frequency in ancestry component 6
• f-pop7: inferred allele frequency in ancestry component 7
• ancestry_component: ancestry component tested by the selection hypothesis model. Note that we have added 1 to the ancestry component numbers to match the terminology used in paper (which orders the components from 1-8 rather than 0-7 for interpretability)
• snp_perc: SV's percentile in the LRS distribution for frequency-matched SNPs
• p_nominal: nominal p-value calculated from the likelihood ratio
• p_adj: adjusted p-value calculated from the likelihood ratio