Bayesian variable selection using an informed reversible jump in imaging genetics: an application to schizophrenia

Modern attempts in providing predictive risk for complex disorders, such as schizophrenia, integrate genetic and brain information in what is known as imaging genetics. In this work, we propose inferential and predictive methods to relate the presence of a complex disorder, schizophrenia, to genetic and imaging features and predict its risk for new individuals. Given functional Magnetic Resonance Image and Single Nucleotide Polymorphisms information of healthy and people diagnosed with schizophrenia, we use a Bayesian probit model to select discriminating variables, while to estimate the predictive risk, the most promising models are combined using a Bayesian model averaging scheme. For these purposes, we propose an informed reversible jump Markov chain Monte Carlo, named data driven or informed reversible jump, which is scalable to high-dimension data when the number of covariates is much larger than the sample size.