WP1 is the consortium’s broad discovery platform and aims to identify markers and predictors of resilience in a large pooled transdiagnostic sample using machine learning, deep learning, and network control theory. It integrates more than 5,500 participants across patient and control cohorts, harmonising outcome-based resilience proxies such as episode recurrence, hospitalisation history, and transitions from high-risk states, alongside trait-based measures, polygenic resilience scores, neuroimaging, immunological, cognitive, and environmental data. The work is organised into four tasks: data harmonisation and proxy definition, data-driven modelling of resilience predictors, network-based analyses of interrelations among predictors, and replication in newly recruited or longitudinal cohorts. A further goal is to test whether these pipelines can function as candidate “digital twin” prototypes for personalised resilience profiling. WP1 is significant because it provides the first large-scale multi-level analysis of resilience and will generate testable markers for stratifying vulnerability and guiding intervention studies.

WP2 uses a hypothesis-driven imaging approach to test two specific neural mechanisms that may underlie individual differences in resilience: a hippocampal network mechanism and a salience network mechanism. These mechanisms will be examined against both outcome-based resilience indicators, such as episode history and psychiatric hospitalisations, and psychological trait resilience measured by questionnaire. Multiple imaging parameters will be used to compare effect sizes and evaluate translational relevance. The large multisite sample of more than 4,000 participants provides strong statistical power and allows the consortium to compare structural and functional markers across modalities. In addition, centile score analyses will assess whether resilience-related variation reflects deviations from normative neurodevelopmental trajectories. WP2 is important because it links resilience to interpretable brain markers that may support risk stratification and future interventions.

WP3 examines the factors that shape outcome-based resilience in people with bipolar disorder, major depressive disorder, and schizophrenia, focusing on the role of enrichment and cognitive/neural flexibility. The central idea is that higher levels of enrichment may buffer the impact of stressful life events on episode recurrence and psychiatric hospitalisation, potentially by enhancing flexibility at both cognitive and neural levels. The work tests whether stressful life events are associated with greater illness recurrence, especially in individuals with low enrichment, and whether enrichment is related to stronger flexibility and better outcomes. It also explores how these variables relate to self-reported psychological trait resilience. WP3 matters because it may identify mechanisms through which environmental enrichment supports resilience and relapse prevention, with potential applications for stratification and intervention design.

WP4 takes a gene-by-environment approach to understand how polygenic risk for autism spectrum disorder and schizophrenia spectrum disorders relates to peer problems and psychotic experiences as a proxy for outcome-based resilience. By analysing both ASD and SSD polygenic risk scores in the same individuals, the work can test additive or interactive genetic effects and identify specific multigenic profiles linked to psychotic experiences or resilience. A second aim is to examine whether peer problems mediate these associations, with high peer problems representing risk and low peer problems representing protection. The project also considers neural factors such as hippocampal connectivity in relation to these pathways. WP4 is significant because it helps clarify how genetic liability and environmental context combine to shape mental health outcomes and points toward interventions focused on social and neural protective factors.

WP5 studies resilience in a uniquely vulnerable cohort exposed to earthquakes, SARS-CoV-2, and forced migration. It aims to identify moderating and mediating factors associated with outcome-based resilience, defined by the presence or absence of psychiatric diagnoses or subthreshold symptoms after these stressors. The work will collect new data to replicate and extend findings from other WPs, especially cognitive and neural resilience mechanisms, while also assessing additional risk and protective factors not covered elsewhere in the consortium. A key strength of WP5 is that it zooms in on real-world, high-stress exposure, allowing the consortium to examine the micro-level dynamics of resilience in a highly relevant setting. WP5 is important because it broadens the generalisability of the consortium’s findings and adds context-specific insights into resilience under severe adversity.

WP6 uses twin modelling to quantify the genetic and environmental contributions to variation in outcome-based resilience, as well as the stability of resilience proxies over time and their protective correlates, including brain-based markers. It also estimates shared genetic and environmental influences between resilience outcomes, brain resilience, and broader protective factors, and tests whether these associations follow causal or non-causal patterns. The consortium expects moderate heritability, substantial unique environmental influence, and meaningful genetic overlap between resilience and protective brain or behavioural features. WP6 is significant because it leverages twin data to address a central methodological challenge in resilience research: disentangling genetic liability from environmental and causal effects. This makes it especially valuable for understanding whether protective factors truly shape resilience or merely co-occur with it.