Random Forests on Distance Matrices for Imaging Genetics Studies
This method addresses the challenge of analyzing non-vectorial brain data like connectivity networks in genetics studies, though it appears incremental as an extension of decision forests.
The authors tackled the problem of detecting genetic variants associated with brain phenotypes in imaging genetics by proposing Random Forests on Distance Matrices (RFDM), a non-parametric regression method that uses distance matrices as responses, and demonstrated its application in an Alzheimer's Disease study with simulation results.
We propose a non-parametric regression methodology, Random Forests on Distance Matrices (RFDM), for detecting genetic variants associated to quantitative phenotypes representing the human brain's structure or function, and obtained using neuroimaging techniques. RFDM, which is an extension of decision forests, requires a distance matrix as response that encodes all pair-wise phenotypic distances in the random sample. We discuss ways to learn such distances directly from the data using manifold learning techniques, and how to define such distances when the phenotypes are non-vectorial objects such as brain connectivity networks. We also describe an extension of RFDM to detect espistatic effects while keeping the computational complexity low. Extensive simulation results and an application to an imaging genetics study of Alzheimer's Disease are presented and discussed.