Xiaoxu received his PhD in Bioinformatics from the Center for Bioinformatics, Peking University, he finished his postdoctoral training at the Department of Neurosciences, University of California, San Diego. He started the study for sperm mosaic mutations causing childhood neurological and psychiatric disorders with wet- and dry-lab methodologies he developed. He further extended his research interests in sperm and brain mosaicism. He developed DeepMosaic, the first Deep-learning-based mosaic variant caller without matched control, and MosaicBase, the first non-cancer mosaic variant knowledge base. Using these technologies he developed, he found sperm clonal mosaic mutations causing life-long transmission risk for normal males. He further used mosaic mutations as developmental markers unrivaled left-right differences and inhibitory neuronal migration patterns in human cortical development.