Lisa

Lisa predicts transcriptional regulators (TRs) associated with differentially expressed or co-expressed gene sets by integrating chromatin accessibility and ChIP-seq data to model regulatory chromatin.

Key Features:

• Integration of Chromatin Data: Integrates histone mark ChIP-seq and chromatin accessibility profiles to construct regulatory chromatin models.
• In Silico Deletion Analysis: Probes chromatin models using TR ChIP-seq peaks or imputed TR binding sites via in silico deletion to identify regulatory elements impacting target genes.
• Imputed TR Cistromes: Incorporates imputed TR binding sites to extend TR cistrome information where direct ChIP-seq data are absent.
• Enhanced Predictive Performance: Demonstrates superior performance in identifying perturbed transcriptional regulators when applied to gene sets from targeted TF perturbation experiments.

Scientific Applications:

• Gene Expression Studies: Identification of key TRs driving differential or co-expression patterns in gene sets.
• Targeted TF Perturbation Experiments: Prioritization of transcription factors perturbed in experiments based on gene set regulatory modeling.

Methodology:

Uses publicly available chromatin accessibility and histone mark ChIP-seq data to build regulatory chromatin models; probes those models with TR ChIP-seq peaks or imputed TR binding sites using in silico deletion; validated by application to targeted TF perturbation gene sets showing improved identification of perturbed TRs.