Bov-ATAC

ATAC-seq peak calling, we follow Encode

1. FRip score, TSS enrichment and Rescue Ratio are used to eliminate the lower quality data
2. Peak called using Macs2

To generate and annotate the peak we borrow the idea developed by Meuleman

1. We perform peak calling for each sample and then use FWHM method to obtain consensus
2. To infer the the biological meaning, we decompose consensus by sample matrix into basis making it is interpretable

In cis-eQTL analysis, we follow GTEX

1. we correct hidden factor by PEER
2. cis-eQTL are detected using QTLtools

Enrichment analysis, we using goshifter developed by Trynka

Estimation of fa

By combining eQTL and ATAC-Seq, we estimate proportion of regulatory variants mapping to ATAC-Seq based on Maximum Likelihood

Data availability

Please check on UCSC genome brower ATAC-seq-UCSC