Back

Integrative genetic and transcriptomic analysis identifies candidate genes and compounds for Fragile X syndrome and autism spectrum disorder

Zhang, W.; Yan, J.; DeGiorgio, M.

2026-10-06 bioinformatics
10.64898/2026.10.03.756403 bioRxiv
Show abstract

Fragile X syndrome (FXS) is the most commonly known monogenic cause of autism spectrum disorder (ASD), yet how FXS biology relates to polygenic architecture of idiopathic ASD remains poorly defined. We integrated publicly available genetic data relevant to ASD and FXS, including population case-control risk, symptom discordance within monozygotic twin pairs, contrasts between ASD and other psychiatric disorders, and genetic variation associated with circulating plasma fragile X messenger ribonucleoprotein (FMRP) abundance. These data were used to identify associated genes and pathways across tissues and prioritize small molecules predicted to reverse genetically associated transcriptional signatures. Cross-tissue analysis identified XRN2 as the strongest association for population ASD risk and PPP3CA as the strongest association with FMRP abundance, providing distinct candidate links to ASD and FXS biology. We found no evidence that population ASD risk and symptom discordance within monozygotic twin pairs share common-variant architecture after accounting for linkage disequilibrium, suggesting that these phenotypes capture at least partially distinct sources of genetic signal. Connectivity mapping identified candidate compounds across eight genetic signatures, including several with biologically plausible mechanisms. Together, these findings provide a genetics-anchored framework for connecting FXS and ASD biology and for prioritizing candidate therapeutic mechanisms and compounds, while highlighting methodological safeguards needed for reliable integrative genomic drug repurposing.

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.