PT  - JOURNAL ARTICLE
AU  - Jian Zhou
AU  - Christopher Y. Park
AU  - Chandra L. Theesfeld
AU  - Yuan Yuan
AU  - Kirsty Sawicka
AU  - Jennifer C. Darnell
AU  - Claudia Scheckel
AU  - John J Fak
AU  - Yoko Tajima
AU  - Robert B. Darnell
AU  - Olga G. Troyanskaya
TI  - Whole-genome deep learning analysis reveals causal role of noncoding mutations in autism
AID  - 10.1101/319681
DP  - 2018 Jan 01
TA  - bioRxiv
PG  - 319681
4099  - http://biorxiv.org/content/early/2018/05/11/319681.short
4100  - http://biorxiv.org/content/early/2018/05/11/319681.full
AB  - We address the challenge of detecting the contribution of noncoding mutation to disease with a deep-learning-based framework that predicts specific regulatory effects and deleterious disease impact of genetic variants. Applying this framework to 1,790 Autism Spectrum Disorder (ASD) simplex families reveals autism disease causality of noncoding mutations by demonstrating that ASD probands harbor transcriptional (TRDs) and post-transcriptional (RRDs) regulation-disrupting mutations of significantly higher functional impact than unaffected siblings. Importantly, we detect this significant noncoding contribution at each level, transcriptional and post-transcriptional, independently. Further analysis suggests involvement of noncoding mutations in synaptic transmission and neuronal development and points to functional relevance of progenitor cell types to ASD. We demonstrate that sequences carrying prioritized proband de novo mutations possess transcriptional regulatory activity and drive expression differentially, and propose a potential link between the quantitative impact of noncoding versus coding mutations in ASD individuals to their IQ. Our predictive genomics framework illuminates the role of noncoding variants in ASD, prioritizes high impact transcriptional and post-transcriptional regulatory mutations for further study, and is broadly applicable to complex human diseases.