Peyton Greenside

Identification and mitigation of pervasive off-target activity in CRISPR-Cas9 screens for essential non-coding elements

The paper identifies off-target effects in CRISPR-Cas9 screens for non-coding elements and shows that strict filtering with specificity scores improves the identification of essential regulatory regions.

Deciphering regulatory DNA sequences and noncoding genetic variants using neural network models of massively parallel reporter assays

The paper introduces MPRA-DragoNN, a CNN model to predict regulatory DNA activity from MPRAs, identifies TF motifs, and explores variant effects in lipid-associated loci.

CrowdVariant: a crowdsourcing approach to classify copy number variants

The paper introduces CrowdVariant, a crowdsourcing framework to classify copy number variants (CNVs) using NGS data, and identifies 1,781 high-confidence CNVs in a reference sample.

Discovering epistatic feature interactions from neural network models of regulatory DNA sequences

The paper introduces DFIM, a method to identify epistatic feature interactions in neural network models of regulatory DNA sequences, validated on simulated and in vivo data.

Impact of regulatory variation across human iPSCs and differentiated cells

The paper examines regulatory variation in iPSCs and differentiated cells, finding lower individual variation in iPSCs and identifying shared chromatin regions for some cell type-specific QTLs.

Evaluation of RNAi and CRISPR technologies by large-scale gene expression profiling in the Connectivity Map

The paper evaluates RNAi and CRISPR technologies by analyzing gene expression profiles in the Connectivity Map, finding that RNAi has strong off-target effects while CRISPR is more accurate.

A Next Generation Connectivity Map: L1000 Platform And The First 1,000,000 Profiles

The paper introduces the L1000 platform, a high-throughput gene expression profiling method, and reports 1.3 million perturbational profiles expanding the Connectivity Map for drug mechanism discovery and functional annotation.

Enhancer connectome in primary human cells reveals target genes of disease-associated DNA elements

This paper uses HiChIP to map enhancer-promoter interactions in primary human cells, linking disease-associated DNA variants to their target genes.

An improved ATAC-seq protocol reduces background and enables interrogation of frozen tissues

Omni-ATAC improves ATAC-seq by reducing background noise and enabling analysis of frozen tissues, revealing disease-associated DNA element activities in human brain structures.

PyBoost: A parallelized Python implementation of 2D boosting with hierarchies

PyBoost and HiBoost are parallelized Python implementations of GeneClass for learning gene regulatory networks from paired transcription factor and motif data, applied to regeneration and differentiation studies.

Reverse-complement parameter sharing improves deep learning models for genomics

Standard genomic CNNs can score a sequence and its reverse complement differently; four weight-sharing layer types remove that inconsistency and are reported to train faster.