A scalable systematic RDoC multi-omics framework to reorganize preclinical psychiatric research beyond diagnostic categories

Psychiatric neuroscience faces a persistent translational gap, largely attributed to the limited construct validity of animal models when interpreted through categorical diagnoses. Influential critiques emphasize that such animal models cannot recapitulate complex psychiatric syndromes, supporting a shift towards dimensional and mechanistic approaches such as the Research Domain Criteria. Despite its conceptual relevance, RDoC has not been systematically applied to integrate large-scale preclinical datasets across multiple levels of analysis.

Here, we present a comprehensive and integrative framework that operationalizes RDoC across one genetic model of depression, the Flinders Sensitive Line. We used a systematic review approach followed by a stratified extraction of data across RDoC domains and units of analysis, enabling structured domain integration of phenotypes and biological mechanisms. To resolve the biological architecture underlying each domain, molecular targets were mapped to KEGG pathways and Gene Ontology identifiers, allowing multi-omics integration of signaling and functional pathways. Pharmacological interventions were classified according to mechanism of action, experimental models were grouped into broader categories, and behavioral outcomes were categorized based on impairment or improvement within each RDoC domain. All analysis were performed with RStudio.

We identified 2.498 records, from which 306 studies met inclusion criteria after removal of duplicates and screening. From these studies we obtained 11.400 experimental comparisons. Integrated analysis revealed that the Flinders Sensitive Line exhibits robust impairments in cognitive and negative valence domains from early developmental stages. Importantly, our analysis demonstrated that the Flinders Resistant Line, often used as the Flinders Sensitive Line control, does not represent a neutral control; relative to Sprague-Dawley rats, it shows enhanced performance in negative and positive valence domains, alongside alterations in arousal and regulatory systems. These findings challenge standard assumptions regarding control conditions in this model. Pharmacological profiling across domains revealed heterogeneous efficacy patterns. For example, selective serotonin reuptake inhibitors showed inconsistent effects within the negative valence domain and limited impact across cognition, social processes, and positive valence systems. In contrast, NMDA receptor antagonists demonstrated more robust improvements in negative valence. Multi-omics pathway analysis integrated with physiological data pointed to several alterations in metabolic and cardiometabolic pathways, as well as disruptions in neurotransmission and synaptic plasticity. Notably, several molecular pathways overlap with those implicated in immune and metabolic disorders, highlighting shared biological substrates. Environmental modulation analysis indicated that both aversive and positive conditions significantly influence domain-specific outcomes at behavioural and molecular levels, reinforcing the context-dependent nature of this model.

This framework represents a systematic strategy to reorganize preclinical evidence beyond disorder-based classifications. By deeply integrating RDoC domains with multi-level biological data, it enables the identification of domain-specific neurobiological signatures and exposes hidden structure within heterogeneous datasets. This approach addresses key limitations in the field and provides a translationally relevant platform for interpreting animal models within a dimensional, domain-level framework.

References

No references