Investigating tissue and cellular dynamics via transcriptomics at a single-cell resolution allows us to identify disease-specific cell states, while mass spectrometry characterizes body fluids at various disease stages. Integration of both modalities from multiple cohorts and samples allows us to study whether quantitative changes in body fluid protein signatures can help explain the cell state changes analyzed by single-cell RNA-seq. The initial aim of this project is to create a consensus atlas of human brain transcriptome in schizophrenia disorder, based on meta-analyses of differential gene expression from multiple single-cell and bulk RNA-seq datasets publicly available. This will serve as a powerful resource to investigators studying neurodevelopmental disorders and can be useful to deconvolute cell type composition in bulk datasets from similar tissues. Further, the atlas will be integrated with high-quality mass spectrometry data from CSF samples of patients (sourced from data available publicly). We will apply machine learning to test whether mRNA expression mapped to proteins can correctly predict the direction of disease in different cohorts. This will tell us whether changes in cellular states in diseased brain tissue can be reflected by protein signatures in CNS fluids such as CSF.