Transcriptomic decoding of surface-based imaging phenotypes and its application to pharmacotranscriptomics

Abstract

Abstract Imaging transcriptomics has become a power tool for linking imaging-derived phenotypes (IDPs) to genomic mechanisms. Yet, its potential for guiding CNS drug discovery remains underexplored. Here, utilizing spatially-dense representations of the human brain transcriptome, we present an analytical framework for the transcriptomic decoding of high-resolution surface-based neuroimaging patterns, and for linking IDPs to the transcriptomic landscape of complex neurotransmission systems in vivo. Leveraging publicly available Positron Emission Tomography (PET) data, we initially validated our approach against molecular targets with a high correspondence between gene expression and protein binding. Subsequently, we used the cortical gene expression profiles of candidate genes to dissect two discrete classes of GABAA-receptor subunits, each characterized by a distinct cortical expression pattern, and to link these to specific behavioural symptoms and traits. Our approach thus represents a future avenue for in vivo pharmacotranscriptomics that may guide the development of targeted pharmacotherapies and personalized interventions.