Inference of Autism Risk Genes Through Comparative Sociogenomics and Molecular Network Analysis

Comparative sociogenomics combines multiple scientific fields to investigate the genetic basis of social behavior across species. Our aim was to uncover the genetic roots of human sociability with possible implications for autism, a neurodevelopmental disorder characterized by social and communication deficits.

We conducted molecular network analysis on 659 sociability-related genes from different animal species, including humans.

We identified a network of 240 genes strongly associated with autism (p < 10^-15), with 194 inferred. These genes were grouped into 23 functional communities related to cell-cell junctions and communication, inflammatory and synaptic signaling, neurotransmitter receptors and semaphorin signaling among the more enriched meta-pathways. Some network genes were clustered in nine chromosomal bands (FDR < 0.25), indicating genes' functional cooperation, shared evolutionary history, and coordinated regulation, and few genes are physically in linkage with ASD genes (within 0.5 cM) or controlled by human-accelerated regions.

The most compelling inferred autism risk genes are MED12, FZD9, and DMD since they are differentially expressed in autistic brains, physically linked to key autism genes, controlled by human-accelerated regions, or mapped to chromosomal regions enriched in network genes. If validated, they could represent novel biomarkers, advancing the understanding of autism's genetic makeup.