A genomic mutational constraint map using variation in 76,156 human genomes

Area of Science:

• Genomics
• Human Genetics
• Evolutionary Biology

Background:

• Purifying natural selection (constraint) is key for studying protein-coding genes in human disorders.
• Assessing constraint in non-protein-coding regions has been challenging.
• Large-scale human genome datasets are crucial for genomic analyses.

Purpose of the Study:

• To create a comprehensive genomic constraint map for the entire human genome.
• To investigate constraint in non-protein-coding regions using a large dataset.
• To improve the identification and interpretation of functional genetic variation.

Main Methods:

• Aggregated and processed data from 76,156 human genomes (gnomAD).
• Developed a refined mutational model incorporating sequence context and genomic features.
• Built a whole-genome constraint map (Gnocchi).

Main Results:

• The average constraint in protein-coding sequences is stronger than in non-coding regions.
• Constrained non-coding regions are enriched for regulatory elements and disease-associated variants.
• A link was found between constrained regulatory elements and constrained protein-coding genes.

Conclusions:

• The Gnocchi map enhances the analysis of non-coding DNA, linking biological annotation, disease association, and natural selection.
• Non-coding constraint can help identify previously unrecognized constrained genes.
• This genome-wide constraint map improves the identification and interpretation of functional human genetic variation.