CSSQ is a R package built to identify differentially bound regions across groups of ChIP-seq samples. It uses robust statistical methods to quantitate, transform and normalize signals using ChIP-seq and control samples before testing them for differential binding. The methods utilized adjust for noise arising due to differences in experimental efficiencies.

Bioconductor page

deepPredict is a framework in progress that is designed to predict gene expression values using epigenetic information. It utilizes multiple deep learning models to process methylation data across the genome to predict gene expression values.

Single cell RNA-seq analysis

• Single cell RNA-seq data from mouse embryonic stem cells that have subtypes of linker histone H1 knocked out were analyzed to identify underlying regulation changes. Data analysis involved scRNA-seq data processing, batch correction, imputation, clustering and differential gene expression.

Allele specific expression analysis

• End-to-end analysis pipeline was built for performing allele specific expression analysis on histone H1 knockout RNA-seq data to identify allelic imbalances caused by the knockout of different histone H1 subtypes.