WP3 Data engineering and statistics to allow integrated analysis of data sets

WP3 will develop a statistical approach for fusion of quantitative modality-specific biological measurements relative to behavioural parameters of social withdrawal, attention, working memory, and sensory processing and link these to behavioral data and clinical assessments. We will use data-driven clustering of individual domain data validated against symptom constellations to identify the highest predicting quantitative domain in patients stratified for low versus high social withdrawal. Moreover, multi-modal data fusion approaches are applied to derive quantitative phenotypes across the different makes (behavioural, genetic, and imaging-derived) and validated against neuropsychiatric measures.