Statistical Methods of Disease-Gene Mapping in Trio-based Next Generation Sequencing

Abstract

Disease-gene mapping plays an important role in improving the development of medical science. As with the development of Next Generation Sequencing technologies, mapping disease genes through rare genetic variants become economic and reliable. De novo mutations as the most extreme form of rare variants played an important role in the occurrence of complex diseases. To detect de novo mutations, case-parent trios are used to perform the sequencing studies. This case-parent design provides the chance to detect disease-causal genes from both de novo mutations and inherited mutations. We proposed three novel methods to map disease genes according to de novo mutation load (fitDNM), allele transmission rate (rvTDT) and compound heterozygous and recessive genes (coreTDT) separately to maximize the statistical power of analysis in case-parent trios. These three methods are then applied to analyze neurodevelopmental/neuropsychiatric disorders. The analysis with fitDNM provides strong statistical evidence supporting two potentially causal genes: SUV420H1 in autism spectral disorder and TRIO in a combined analysis of the four neurodevelopmental/neuropsychiatric disorders investigated. The application of rvTDT on epileptic encephalopathy (EE) trios find that dominant (or additive) inherited rare variants are unlikely to play a substantial role within EE genes previously identified through de novo mutation studies.