Functional networks of the brain underlie cognitive and emotional processing and are impacted by neurodevelopment and modulated by neurotransmitters. To better understand the neurobiology and psychology of cognitive and emotional processing, it is crucial to combine functional and molecular neuroimaging methods to examine both neural activity/networks and neurotransmitter systems. However, this combination is rare and current methods used to combine multimodal neuroimaging data with other data modalities are suboptimal as they almost exclusively first estimate parameters for each individual and modality separately and then combine them in the second step to assess their relations. This does not take into account the inherent similarities between individuals and between data modalities within individuals and introduces bias in estimates and reduces statistical power. In this project, we aim to develop more robust methods for combining multimodal neuroimaging data with physiological and behavioral data by joint estimation of parameters and relations between parameters in the whole sample simultaneously. We will initially employ a Bayesian hierarchical modeling framework, but also look at more advanced artificial intelligence methods.