Epigenome-augmented eQTL-hotspots reveal genome-wide transcriptional programs in 36 human tissues

Area of Science:

• Genomics
• Systems Biology
• Computational Biology

Background:

• Expression quantitative trait loci (eQTLs) are crucial for understanding transcriptional regulation.
• Identifying eQTLs genome-wide faces limitations due to stringent false discovery rate controls.
• Discovering tissue-specific regulatory mechanisms requires advanced analytical methods.

Purpose of the Study:

• To develop a novel method for identifying eQTL-hotspots, defined as regions with frequent, multiple eQTL associations.
• To create a machine-learning model (E-SpotFinder) for augmented discovery of tissue- or cell-type-specific eQTL-hotspots.
• To reconstruct comprehensive genome-wide cis-regulatory networks and identify transcriptional programs.

Main Methods:

• Utilized a non-homogeneous Poisson process to identify 125,489 eQTL-hotspots from public data across 59 human tissues/cell types.
• Stratified eQTL-hotspots based on sequence and epigenomic characteristics.
• Developed and applied the E-SpotFinder machine-learning model to 36 tissues/cell types for enhanced eQTL-hotspot discovery.

Main Results:

• Identified 125,489 eQTL-hotspots and classified them into two distinct groups.
• Augmented discovery yielded 655,402 eSNPs and reconstructed a network of 2,725,380 cis-interactions.
• Identified 52,012 modules representing unique transcriptional programs.

Conclusions:

• The study presents a framework for epigenome-augmented eQTL analysis.
• Comprehensive genome-wide cis-regulatory networks were constructed across diverse human tissues.
• The findings provide new insights into the mechanisms of transcriptional regulation and functional genomics.