About

Robbee Wedow is an assistant professor of sociology and data science at Purdue University. His main research interests are in statistical and computational genetics, behavioral genetics, and gene-environment interactions. He uses recent advances in genetic data collection, biobank-scale data, and methodological developments in statistical genetics to carry out his research. Dr. Wedow completed his Ph.D. at the University of Colorado Boulder in 2018 and completed his postdoctoral training in 2022 at the Stanley Center for Psychiatric Research at the Broad Institute of MIT and Harvard, the Analytic and Translational Genetics Unit at Massachusetts General Hospital, and the Department of Epidemiology at the Harvard T.H. Chan School of Public Health.

Research topics

• Computer Science
• Genetics
• Biology
• Medicine
• Psychiatry
• Internal medicine
• Psychology
• Computational biology

Selected publications

• Illuminating the complex interplay of risk factors for depression symptoms within a large-scale US longitudinal cohort

  Social Psychiatry and Psychiatric Epidemiology · 2026-05-05

  articleOpen accessSenior author

  Understanding depression susceptibility requires research to assimilate a vast and complex array of risk factors comprehensively. We quantify the influence of prominent individual-level and wider social environmental risk factors on symptom occurrence. We leverage data from the National Longitudinal Study of Adolescent to Adult Health (US) to integrate factors spanning biological (genetic), psychological (health, personality, positive cognition), social (family, peers, school, and neighbourhood), built (geocoded healthcare, education, religion, crime, poverty, political climate), and natural (geocoded population density, rainfall, and urbanicity) systems. We test their contribution to concurrent adolescent depression symptoms (W1), one year later (W2), and a decade later in early adulthood (W4) using relatedness-based linear mixed models (European-like ancestries only; N = 3,867). First, a subset of 81 individual-level factors together explained 85.5% of the variance in concurrent adolescent depression symptoms, 68.4% one year later, and 52.5% in adulthood. When split by domains, positive cognition (feeling accepted and loved) contributed most in adolescence (W1 15.1% [SE = 4%]; W2 6.6% [2%]), and domains contributed equally in adulthood. Second, all 162 factors capturing interconnected genetic, psychological, social, built, and natural systems explained 80.5% in concurrent depression symptoms, 54.7% one year later, and 44.9% in adulthood. The psychological system (W1 38.5% [4%]; W2 21.9% [3%]; W4 7.7% [1%], psychological-genetic interaction (W1 26.7% [4%]; W2 25.1% [4%]; W4 12.4% [3%]), social-genetic interaction (W1 10.1% [3%]; W2 11.8% [3%]; W4 9.6% [3%]) explained most variance in depression symptoms at all ages. We provide a holistic understanding of depression risk, where feeling supported and accepted was most crucial, and emphasise the complexity of modelling the environment.

  PublisherDOI

• Cellular morphology emerges from polygenic, distributed transcriptional variation

  bioRxiv (Cold Spring Harbor Laboratory) · 2026-03-13

  articleOpen access

  Abstract Height and most disease risk are known polygenic traits: characteristics governed by multiple genes at different loci instead of a select few. Though we are beginning to understand how genetic variation impacts cell morphology, whether such an analogous polygenic architecture operates at the cellular level, where morphology integrates cytoskeletal organization, organelle positioning, and metabolic state, has yet to be systematically tested. Here, we demonstrate that cellular morphology behaves as a polygenic trait by integrating multimodal modeling, perturbation profiling, and population-scale genetic variation. A shared latent-space autoencoder trained on four large-scale perturbation datasets predicts morphology from gene expression and generalizes without retraining to matched RNA-seq and Cell Painting profiles from 100 genetically diverse iPSC donors. The model predicted 17 morphological features (R² > 0.6, permutation FDR q < 0.05), enriched for spatial organelle distribution and cytoskeletal architecture. Predictive performance does not arise from dominant gene–phenotype relationships: individual genes contribute modestly, and marginal gene–morphology correlations are uniformly weak, revealing a distributed regulatory architecture. Despite this polygenicity, CRISPR perturbation data from the JUMP consortium validates specific model-prioritized genes, such as the cytoskeletal regulator TIAM1 , membrane trafficking factor RAB31 , and mitochondrial-associated membrane transporter ABCC5 , as molecular anchors whose disruption produces feature-specific morphological shifts. Transcriptome-wide association analyses identify correlational variant–gene–morphology chains linking cis-regulatory variation through mitochondrial metabolism ( PDHX ) and iron transport ( SLC11A2 ) to cellular architecture. These results establish cellular morphology as a polygenic systems phenotype, extending the omnigenic framework to the cellular level and providing a biological basis for interpreting cross-modal prediction in functional genomics.

  PublisherOA PDFDOI

• Social connection can mitigate depression risk

  PsyArXiv (OSF Preprints) · 2026-03-10

  preprintOpen access

  Enhancing social connection is a widely implementable target to reduce depression risk. Yet, little is known about whether social connection can meaningfully offset more immutable forms of risk. We use causal inference to estimate how disparities in depression by sex, race, family socioeconomic position, and genetic liability can be reduced following a hypothetical social connection intervention. In a U.S. representative sample of ~10,000 people, we demonstrate that social connection mitigates sex, genetic, and socioeconomic disparities in depression symptoms, which persisted over a decade later. However, reductions were not consistent when stratified by sex and racial groups. These observational data showcase how social paradigms could mitigate even genetic and structural risks, and highlight the need for intersectionally informed mental health interventions and policy.

  Publisher

• Social connection can mitigate depression risk

  2026-03-10

  articleOpen accessSenior author

  Enhancing social connection is a widely implementable target to reduce depression risk. Yet,little is known about whether social connection can meaningfully offset more immutableforms of risk. We use causal inference to estimate how disparities in depression by sex,race, family socioeconomic position, and genetic liability can be reduced following ahypothetical social connection intervention. In a U.S. representative sample of ~10,000people, we demonstrate that social connection mitigates sex, genetic, and socioeconomicdisparities in depression symptoms, which persisted over a decade later. However,reductions were not consistent when stratified by sex and racial groups. These observationaldata showcase how social paradigms could mitigate even genetic and structural risks, andhighlight the need for intersectionally informed mental health interventions and policy.

  PublisherDOI

• Heritability

  The Wiley Blackwell Encyclopedia of Health, Illness, Behavior, and Society · 2025-11-03

  other1st authorCorresponding

  Abstract Heritability is a statistical parameter that conveys how much variance in an outcome can be attributed to genetic variance. Sometimes estimated as broad sense heritability from twin and family models and sometimes as narrow sense heritability from the additive effects from genome‐wide association studies, heritability is expressed as the ratio of genetic variance to phenotypic variance and ranges from 0 to 1. As a statistical concept, heritability is not immutable or fate; rather it is population specific and context dependent and can change if the environment also changes.

  PublisherDOI

• M40. PREDICTING DEPRESSION FROM ADOLESCENCE TO ADULTHOOD USING MACHINE LEARNING: INSIGHTS FROM A NATIONALLY REPRESENTATIVE COHORT

  European Neuropsychopharmacology · 2025-10-01

  articleOpen access

  Early identification of individuals at risk for depression is essential for timely intervention. While genetic risk scores are increasingly used in psychiatric research, their predictive utility remains limited. This study assessed whether adolescent psychosocial and environmental data could predict depressive outcomes up to two decades later. We further evaluated sex-specific model performance and the incremental value of polygenic scores (PGS) for depression. Data were drawn from Wave I of the National Longitudinal Study of Adolescent to Adult Health (Add Health; N = 15,701). Six machine learning (ML) algorithms—logistic regression, support vector machines (linear and RBF kernels), random forest, XGBoost, and a fully connected neural network—were trained on adolescent psychosocial and environmental features. Outcomes included depressive symptoms in adolescence and adulthood, as well as clinical diagnoses of depression at both timepoints. Models were evaluated using the area under the receiver operating characteristic curve (ROC-AUC). Analyses were stratified by sex and repeated with the inclusion of depression PGS to assess their contribution to predictive performance. We also performed feature importance to verify the most relevant questions. XGBoost consistently yielded the highest predictive performance across outcomes. For adolescent depressive symptoms, the top AUC was 0.849 (females). Predictive accuracy declined for adult depressive symptoms (maximum AUC = 0.661) and adult clinical diagnoses (AUC = 0.636), likely reflecting increased influence of life-course exposures. Models trained on adolescent data predicted clinical diagnoses in adolescence with moderate accuracy (AUC = 0.788). Non-linear models consistently outperformed linear approaches, but linear models can still be used with some success. Feature importance analyses indicated shifts over time: emotional well-being, parental connectedness, and academic behaviors were significant predictors during adolescence, whereas general health, behavioral risks, and family history were more relevant to adult outcomes. Sex-stratified models revealed distinct predictive profiles. The addition of PGS did not significantly improve the prediction when used in combination with the environment. Models with and without PGS yielded nearly identical AUCs, suggesting that the current genetic risk scores have limited incremental utility in the presence of rich environmental data. Adolescent psychosocial and environmental variables allow for moderate-to-high prediction of short- and long-term depressive outcomes, particularly among females. While predictive performance declines over time, early-life features remain informative for adult diagnoses. The lack of improvement with PGS suggests that the added predictive value of current genetic predictors may be limited in the presence of comprehensive environmental and psychosocial data. These findings support adolescence as a critical window for identifying individuals at elevated risk for depression and suggest that ML models, especially XGBoost, offer promising tools for early risk stratification in psychiatric research.

  PublisherDOI

• What machine learning teaches us about depression prediction across the life course: An exploratory comparison of predictive models

  SSM - Population Health · 2025-11-20

  articleOpen accessSenior author

  Identifying individuals at risk for depression early is important for preventing long-term mental health issues. However, the variability in depression severity, duration, and triggers complicates predictions. This study explores whether machine learning models can outperform traditional methods, like Logistic Regression, in predicting self-reported depressive symptoms and clinical depression during adolescence and adulthood. We applied five machine learning models with varying complexity levels — Logistic Regression, Random Forest, XGBoost, Support Vector Machine, and Neural Networks — using data from a nationally representative longitudinal study of the U.S., which tracked participants for 20 years. The models were trained with early-life predictors (ages 12–18) from Wave I, including environmental factors (family, school, health) and genetic predispositions (polygenic scores) from Wave IV. Models were evaluated on their ability to predict depressive symptoms and clinical diagnoses in both adolescence and adulthood. After evaluating the performance of all five models, XGBoost emerged as the most effective, with a 0.02 increase in ROC-AUC compared to the benchmark Logistic Regression model. While this is a slight performance improvement, overall, Logistic Regression performs about as well as many of our ML models. Early-life data showed strong predictive value for depressive symptoms and clinical diagnoses in adolescence and adulthood, highlighting adolescence as a critical period. Polygenic scores added minimal predictive power when combined with environmental data. Feature importance analyses identified self-perception and physical health as key predictors of depressive symptoms, while trauma and life-changing events were more influential for clinical depression. • Logistic regression performs similarly or better than 4 ML models in predicting depression. • XGBoost outperformed logistic regression, though the improvements were marginal. • XGBoost identified depression risk factors that are consistent with previous literature. • Trauma and life-changing events were the strongest predictors of clinical depression. • Polygenic scores did not improve depression prediction across any model.

  PublisherDOI

• Gene–Environment Interaction

  The Wiley Blackwell Encyclopedia of Health, Illness, Behavior, and Society · 2025-11-03

  other1st authorCorresponding

  Abstract Gene–environment interactions (GxE) posit an interactive role between genetics and the environment in which genes influence sensitivity to environmental exposures. GxE offers insight into one specific type of gene–environment interplay, since this framework demonstrates that genetic effects on traits or behaviors are contingent upon environmental factors. Gene–environment interaction effect may operate through biological, behavioral, or epigenetic pathways, and they are distinct from gene–environment correlation, in which genes influence exposure to environments. Though gene–environment interactions are theoretically an important component of understanding human behavior and disease risk, these studies are challenged by the truly enormous sample sizes necessary to statistically identify these effects.

  PublisherDOI

• Genome-wide association study of adolescent-onset depression

  medRxiv · 2025-09-26 · 1 citations

  preprintOpen access

  Adolescent depression is a heritable psychiatric condition with rising global prevalence and severe long-term outcomes, yet its biological underpinnings remain poorly understood. We conducted the first genome-wide association study of adolescent-onset depression, comprising 102,428 cases (diagnosis or clinical symptom thresholds) and 286,911 controls, including diverse ancestries. Cross-ancestry meta-analysis identified 52 independent variants across 17 loci; European-only analysis found 61 variants at 29 loci, with a SNP-based heritability of 9.8%. Comparative analyses revealed two genes unique to adolescent-onset versus lifetime depression, enriched in neuronal subtypes, and two genes as potential drug repurposing targets. Polygenic scores were associated with adolescent-onset depression across ancestries, persistent depression trajectories, more severe outcomes, as well as reduced cortical volume, surface area and white matter integrity. Genetic correlation and Mendelian randomisation analyses support shared genetic liability and causal links with early puberty and modifiable health and behavioural risk factors. These findings uncover novel genetic loci and refine biological pathways underlying adolescent-onset depression, revealing age-specific mechanisms and early intervention opportunities.

  PublisherOA PDFDOI

• Gene–Environment Correlation

  The Wiley Blackwell Encyclopedia of Health, Illness, Behavior, and Society · 2025-11-03

  other1st authorCorresponding

  Abstract Gene–environment correlation (rGE) is a major theoretical and empirical lens that explores the relationship between genetics and the environment. Gene–environment correlation operates through three major theoretical types – passive, evocative, and active. Gene–environment correlation, in which genes influence exposure to environments, is distinct from gene–environment interactions, in which genes influence sensitivity to environmental exposures.

  PublisherDOI

Frequent coauthors

• Raymond K. Walters

  Broad Institute

  99 shared

• Henry R. Kranzler

  Washington University in St. Louis

  94 shared

• Benjamin M. Neale

  Massachusetts General Hospital

  76 shared

• Aarno Palotie

  Institute for Molecular Medicine Finland

  64 shared

• Renato Polimanti

  52 shared

• Josef Frank

  Heidelberg University

  51 shared

• Howard J. Edenberg

  Indiana University – Purdue University Indianapolis

  51 shared

• Leah Wetherill

  University School

  51 shared

Education

• PhD , Sociology; Institute for Behavioral Genetics; Institute of Behavioral Science

  University of Colorado Boulder

• BS Biochemistry and Molecular Biology; BA Sociology, Chemistry and Biochemistry; Biology; Sociology

  University of Notre Dame

  2011