System-level identification of cellular functions underlying human's high-order brain function based on functional module-centric transcriptome analysis

Abstract

Identification of molecular mechanisms underlying human’s high-order brain functions and its dysfunction is a fundamental component of successful drug development for related diseases. Despite many neurobiological studies, yet molecular mechanisms are largely unknown to comprehensively account for the complex phenotypes especially in human. Although advent of bioinformatics analysis based on transcriptome data enables us to efficiently discover candidate molecular components, human brain analysis may require additional sensitivity to the tran-scriptomic change of cellular functions for the following reasons. First, the high-order brain functions are enormously complex, which makes it necessary to decipher system-level mechanisms rather than few single molecules. Second, the type and extent of dysfunction in, for example cognition, is highly heterogeneous between individuals even with the same disease and in vivo human brain transcriptome data is relatively rare due to the difficulty of sampling and quality control. Considering these, I performed two inter-related researches in two different directions to identify the extensive cellular functions that are associated with human’s high-order brain functions and dysfunction by analyzing two different human brain transcriptome data with a sensitivity-enhanced bioinformatics approach. First, I carried out a comparative transcriptomic analysis of human and chimpanzee cerebral cortex to identify extensive cellular functions uniquely associated with human’s pronounced pheno-typic characteristics such as enhanced cognition and vulnerability to neurodegenerative diseases and psychiatric disorders. Differently from previous studies, I measured activity change of functional modules rather than individual genes by analyzing change of modular gene expression and of intra-module coexpression between species. I showed that the intra-module coexpression change analysis most extensively identified changes of functional modules c...