BRAIN-MAGNET: A functional genomics atlas for interpretation of non-coding variants

Area of Science:

• Genomics
• Neuroscience
• Computational Biology

Background:

• Interpreting non-coding genetic variants for diseases is challenging.
• Non-coding regulatory elements (NCREs) play crucial roles in gene regulation.
• Understanding NCREs is vital for deciphering genetic contributions to neurological disorders.

Purpose of the Study:

• To functionally annotate NCREs in human brain development using ChIP-STARR-seq.
• To develop BRAIN-MAGNET, an AI tool for predicting NCRE activity and identifying critical nucleotides.
• To leverage the NCRE atlas and BRAIN-MAGNET for fine-mapping GWAS loci and prioritizing rare variants in neurogenetic disorders.

Main Methods:

• Chromatin immunoprecipitation coupled to self-transcribing active regulatory region sequencing (ChIP-STARR-seq) for NCRE functional annotation.
• Development and validation of BRAIN-MAGNET, a convolutional neural network for NCRE activity prediction.
• Application of the functional genomics atlas and BRAIN-MAGNET to analyze common neurological traits and rare neurogenetic disorders.

Main Results:

• Generated a functional genomics atlas of NCREs in human brain development, revealing insights into neural stem cells and early NCRE priming.
• BRAIN-MAGNET accurately predicts NCRE activity and identifies essential nucleotides from DNA sequence.
• Demonstrated the utility of BRAIN-MAGNET in fine-mapping GWAS loci and prioritizing candidate disease-causing variants.

Conclusions:

• The NCRE atlas and BRAIN-MAGNET provide a powerful resource for interpreting non-coding genetic variation.
• This approach can aid in identifying enhanceropathies and understanding the genetic basis of neurological disorders.
• Integrating functional genomics and AI facilitates the interpretation of complex genetic data in human diseases.