Michaelson, J. J. | Search bioRxiv

bioRxiv2018

Calculating the Effects of Autism Risk Gene Variants on Dysfunction of Biological Processes Identifies Clinically-Useful Information

Autism spectrum disorders (ASD) are neurodevelopmental conditions that are influenced by genetic factors and encompass a wide-range and severity of symptoms. The details of how genetic variation contributes to variable symptomatology are unclear, creating a major challenge for translating vast amounts of data into clinically-useful information. To determine if variation in ASD risk genes correlates with symptomatology differences among individuals with ASD, thus informing treatment, we developed an approach to calculate the likelihood of genetic dysfunction in Gene Ontology-defined biological processes that have significant overrepresentation of known risk genes. Using whole-exome sequence data from 2,381 individuals with ASD included in the Simons Simplex Collection, we identified likely damaging variants and conducted a clustering analysis to define subgroups based on scores reflecting genetic dysfunction in each process of interest to ASD etiology. Dysfunction in cognition-related genes distinguished a distinct subset of individuals with increased social deficits, lower IQs, and reduced adaptive behaviors when compared to individuals with no evidence of cognition-related gene dysfunction. In particular, a stop-gain variant in the pharmacogene encoding cycloxygenase-2 was associated with having an IQ<70 (i.e. intellectual disability), a key comorbidity in ASD. We expect that screening genes involved in cognition for deleterious variants in ASD cases may be useful for identifying clinically-informative factors that should be prioritized for functional follow-up. This has implications in designing more comprehensive genetic testing panels and may help provide the basis for more informed treatment in ASD.

genetics↗

bioRxiv2017

TiSAn: Tissue Specific Variant Annotation

Measures of general deleteriousness, like CADD or PolyPhen, have become indispensable tools in the interpretation of genetic variants. However, these measures say little about where in the organism these deleterious effects will be most apparent. An additional, complementary measure is needed to link deleterious variants (as determined by e.g., CADD) to tissues in which their effect will be most meaningful. Here, we introduce TiSAn (Tissue Specific Annotation), a tool that predicts how related a genomic position is to a given tissue (http://github.com/kevinVervier/TiSAn). TiSAn uses machine learning on genome-scale, tissue-specific data to discriminate variants relevant to a tissue from those having no bearing on the development or function of that tissue. Predictions are then made genome-wide, and these scores can then be used to contextualize and filter variants of interest in whole genome sequencing or genome wide association studies (GWAS). We demonstrate the accuracy and versatility of TiSAn by introducing predictive models for human heart and human brain, and detecting tissue-relevant variations in large cohorts for autism spectrum disorder (TiSAn-brain) and coronary artery disease (TiSAn-heart). We find that TiSAn is better able to prioritize genetic variants according to their tissue-specific action than the current state of the art method, GenoSkyLine.

bioinformatics↗