BART Cancer

BART Cancer predicts transcriptional regulators driving gene expression changes across 15 cancer types from TCGA using the Binding Analysis for Regulation of Transcription (BART) methodology.

Key Features:

• Dataset integration: Integrates over 10,000 TCGA RNA-seq gene expression profiles with more than 7,000 ChIP-seq datasets from the Cistrome Data Browser and GEO and incorporates computational predictions from the Cistrome Cancer database.
• Cancer coverage: Analyzes transcriptional regulation across 15 distinct cancer types using primary cancer samples from TCGA.
• Regulatory scope: Predicts activities of over 900 transcriptional regulators, including transcription factors and chromatin regulators, implicated in upregulation and downregulation of genes.
• Predictive algorithm: Employs the BART algorithm (Binding Analysis for Regulation of Transcription) to infer putative transcriptional regulators from gene expression signatures.
• Differential expression basis: Generates predictions based on differential gene expression patterns observed between cancerous and normal tissue samples.
• Regulatory profiles: Produces genomic profiles of active transcriptional regulators derived from primary tumor expression data and ChIP-seq evidence.

Scientific Applications:

• Transcriptional driver identification: Identification of transcriptional regulators responsible for differentially expressed genes in specific cancer types.
• Therapeutic target prioritization: Prioritization of candidate regulatory factors and chromatin regulators as potential therapeutic targets or biomarkers.
• Epigenetic and regulatory characterization: Characterization of epigenetic and transcriptional regulatory landscapes underlying oncogenic gene expression programs.
• Comparative cancer analysis: Comparative analysis of regulator activities across multiple cancer types using integrated RNA-seq and ChIP-seq data.

Methodology:

Integrates TCGA RNA-seq profiles with ChIP-seq datasets from the Cistrome Data Browser and GEO, incorporates Cistrome Cancer database predictions, and applies the BART algorithm to predict transcriptional regulators from differential gene expression between cancer and normal samples.