The Mallard Lab — Active Work
Our work sits at the crossroads of statistical genetics, psychiatric genomics, and computational neuroscience. Across projects, we aim to explain why mental disorders so often co-occur, why they can look so different across individuals, and how genetic risk translates into biological and psychological outcomes.
Research Theme 01
Externalizing disorders — including ADHD, conduct disorder, antisocial personality disorder, and substance use disorders — are among the most heritable and most consequential conditions in psychiatry. They account for enormous individual and societal burden, yet their shared genetic foundations remain incompletely understood. A central challenge is that these disorders rarely occur in isolation: they cluster together in individuals and families, suggesting they arise from a common etiology.
Our lab uses large-scale multivariate genomic methods to characterize how genetic liability is shared and differentiated across the externalizing spectrum. Rather than studying disorders one at a time, we model their overlapping structure directly — identifying both the shared genetic architecture that explains comorbidity and the disorder-specific signals that account for why individuals with similar risk profiles nonetheless present differently.
The lab is centrally involved in the Externalizing Consortium, PsycheMERGE Network, and Psychiatric Genomics Consortium, where we lead and contribute to some of the largest genomic studies of externalizing psychopathology to date. Recently, the lab co-led the Externalizing Consortium’s flagship discovery paper, which analyzed data from more than four million individuals and generated polygenic tools with substantial predictive utility, while also yielding new hypotheses about the synaptic and neurodevelopmental pathways through which externalizing risk operates.
Representative Publications
2026 — medRxiv
Genomic insights into substance use and disinhibitory disorders
By analyzing data from over four million individuals, this study mapped the shared genetics of externalizing spectrum disorders at unprecedented scale, identifying more than 900 putative causal genes and pinpointing key synaptic pathways. It also introduced a polygenic score explaining approximately 12% of variance in externalizing outcomes — one of the strongest predictive instruments yet developed for these outcomes.
2026 — Nature
Mapping the genetic landscape across 14 psychiatric disorders
This large-scale cross-disorder study mapped shared and disorder-specific genetic influences across 14 psychiatric disorders, identifying five broad genomic factors that captured much of their overlapping liability. Particularly relevant to our research, the paper modeled a substance use disorders factor spanning alcohol, cannabis, nicotine, and opioid-related phenotypes, and modeled its relationships with other transdiagnostic factors.
2024 — Nature Mental Health
Characterizing the phenotypic and genetic structure of psychopathology in UK Biobank
Analyzing genetic, brain imaging, and clinical data from over 400,000 individuals, this study showed that domain-level transdiagnostic factors are significantly more informative for genomic discovery than a single general psychopathology factor — challenging the theoretical utility of an all-encompassing p-factor for understanding psychiatric comorbidity and the unique biological correlates it tends to obscure.
Research Theme 02
Mental disorders are not simply adult conditions with adult causes — they emerge from developmental processes that unfold across the lifespan, and often have their roots in early neurodevelopment. Understanding when and where in the brain genetic risk manifests is important to explaining how psychiatric disorders arise and to identifying potential windows for prevention and intervention.
A second line of work in the lab examines how liability for externalizing psychopathology relates to the structure and function of the brain. This includes imaging genetics studies of cortical development and organization, as well as projects that use spatiotemporal transcriptomic data to ask when and in which cell types psychiatric risk genes are expressed across neurodevelopment. We view these as complementary approaches: imaging genetics can reveal where in the brain genetic risk manifests; neurogenomic annotation can illuminate the molecular mechanisms through which it operates.
Current projects in this area focus on the neurodevelopmental pathways linking externalizing liability to altered cortical organization and other neurodevelopmental correlates. This work draws on methods from statistical genetics, genomics, and brain imaging — areas that together can address questions that are difficult to approach from within any single discipline.
Representative Publications
2025 — American Journal of Psychiatry
Characterizing the pleiotropic architecture of impulsivity and its links to psychopathology and neurodevelopment
Using genetic data from over 130,000 individuals, this study showed that specific facets of impulsivity are driven by distinct neurodevelopmental pathways — arguing against treating all impulsive traits as interchangeable. Sensation seeking and lack of perseverance, for instance, exhibit largely unique biological correlates despite their surface similarity, with implications for how impulsivity is conceptualized in psychiatric genetics and neuroscience.
2023 — Nature Communications
Multivariate genomic architecture of cortical thickness and surface area at multiple levels of analysis
This study modeled the shared genetic architecture of cortical thickness and surface area across the cortex, identifying broad genomic factors that capture how brain regions cluster together genetically. These patterns aligned with biologically and functionally meaningful features of cortical organization and showed that genetic influences on surface area were linked to general cognitive function, with more limited associations observed across most psychiatric disorders.
2021 — Nature Neuroscience
X-chromosome influences on neuroanatomical variation in humans
This study demonstrated that common X-linked genetic variation contributes meaningfully to individual differences in human brain structure, with effects that differ between males and females. The work highlights the importance of sex chromosome variation — typically excluded from standard GWAS — for understanding neuroanatomical diversity and its potential links to sex differences in psychiatric risk.
Research Theme 03
Progress in psychiatric genomics depends not only on larger samples, but on better analytic tools. Standard genome-wide association studies analyze one trait at a time, which is a poor fit for a field defined by comorbidity, heterogeneity, and multidimensional outcomes. Developing methods that allow researchers to model the genetic architecture of multiple traits jointly — improving statistical power, biological interpretability, and inferential precision — is a growing area of focus for the lab.
Lab members have contributed to the development and extension of Genomic SEM (Genomic Structural Equation Modeling), a widely used framework for fitting structural models to GWAS summary statistics across multiple traits. Specific contributions include work on stratified Genomic SEM, genomic structural invariance testing, and applications of these tools to psychiatric comorbidity and neurodevelopment. The lab continues to refine and apply these approaches, and is actively working on new methods to better understand pleiotropic genetic architectures.
Representative Publications
2025 — Nature Genetics
Comparison of the multivariate genetic architecture of eight major psychiatric disorders across sex
Applying multivariate genomic methods to GWAS data from eight major psychiatric disorders stratified by sex, this study characterizes how the genetic architecture of psychiatric comorbidity differs between males and females — revealing meaningful sex differences in the structure of shared genetic liability that are obscured when sexes are combined in standard analyses.
2025 — medRxiv
Genomic taxometric analysis of negative emotionality and major depressive disorder highlights a gradient of genetic differentiation across the severity spectrum
This paper introduces genomic taxometric analysis (GTACCC), a new method for evaluating the latent structure of psychiatric phenotypes at the genomic level. Applied to negative emotionality and major depressive disorder, the approach reveals that genetic differentiation across the severity spectrum is better characterized as a gradient than as a categorical boundary — with implications for how psychiatric dimensions are defined and studied.
2022 — Nature Genetics
Genetic architecture of 11 major psychiatric disorders at biobehavioral, functional genomic and molecular genetic levels of analysis
Using Genomic SEM to jointly model GWAS data from eleven major psychiatric disorders, this study characterizes shared genetic architecture across biobehavioral, functional genomic, and molecular levels of analysis — identifying common and disorder-specific pathways that cut across diagnostic boundaries and pointing toward new targets for mechanistic research.
Connect
We welcome inquiries from prospective trainees, scientific collaborators, journalists, and others interested in the lab’s research. If you would like to learn more about our work or explore opportunities to connect, please reach out.
Get in Touch