Core C

Core C: ABSTRACT Core C will support Projects 1-3 in their data analysis, ensuring that these analyses are conducted by rigorous methods and making the data interpretable and reproducible. Core C will pay attention to data types and statistical power, while taking into account important confounding factors that are present in the human population. Core C will, in addition, implement a new method called Shifting Pair Analysis (SPA), based on k- Top Scoring Pairs (kTSP) that was originally invented by Dr. Geman (a PI of this Core) in 2004. Core C MPIs have made fundamental contributions to computational biology (e.g., the introduction of Gibbs sampler and random forests). Taking advantage of their expertise, Core C will provide standardized analytic pipelines for general data analysis, such as power analysis and sample size estimation, group comparison, correlation analysis, and other statistical analyses (Aim 1). Furthermore, Core C will provide customized analytic pipelines and guidance for molecular data analysis (Aim 2), and neuroimaging and multimodal data analysis (Aim 3). Core C will also provide guidelines for controlling confounding factors and biological variables in molecular, neuroimaging, and multimodal data analyses. Core C will assist Projects 1 and 3 in investigating pathological changes in the olfactory epithelium (OE), particularly immune/inflammatory changes and redox imbalance, across humans and mice through immunohistochemistry, single-cell RNA-sequencing, and bulk RNA- sequencing. To study the impact of the OE pathology on clinical manifestations in schizophrenia (SZ) via the olfactory-prefrontal circuits, Core C will guide Project 3 to identify key molecules for the OE pathology in SZ, sub-stratify SZ patients, and perform causal mediation analysis to study relationships among OE pathology, olfactory-prefrontal circuits, and clinical manifestation. Core C will also guide Project 1 to analyze mouse ex vivo brain imaging data, assist Project 2 in analyzing mouse electrophysiological data and comparing the connectivity of the olfactory-prefrontal circuits between control mice and human healthy subjects. In addition, Core C will guide Project 3 to conduct a pilot longitudinal study. In summary, Core C will support Projects by central guidance in data analysis using conventional methods in psychiatry and other medical research (Aims 1-3). The other important and innovative role of Core C is to implement a new method (SPA) (Aim 4). This is a modified version of kTSP, which has been used in cancer and other disease research, in particular enormously contributing to biomarker exploration. But this method has not yet been applied to psychiatry. SPA is advantageous in prioritizing key molecules in samples with high heterogeneity via analyzing relationships between variables, which is less affected by confounding factors. SPA will be used in Projects as an explorative approach to validate the results in a more sensitive manner. Together, Core C will provide integrative data science support with both solid/conventional and innovative/novel analytical pipelines.