Abraham Nunes MD, PhD, MBA, FRCPC Assistant Professor and Mood Disorders Psychiatrist Department of Psychiatry, Dalhousie University 5909 Veterans Memorial Lane (Room 3083B) Abbie J. Lane Memorial Building QE I.I. Health Sciences Centre Halifax, Nova Scotia, Canada, B3H 2E2 E-mail: mylastname at dal dot ca (ACADEMIC MATTERS ONLY) NOTE: I will not respond to clinical matters sent to this email address.

About Me

My research falls broadly within the field of computational psychiatry, where I use both theory-driven (computational modeling of psychiatric conditions) and theory-free (advanced statistical analysis and machine learning) approaches to study heterogeneity and the computational pathophysiology of bipolar disorder (BD). To this end, I work with synthetic, genomic, neuroimaging, neuropsychological, behavioural, clinical, and interventional data. My mathematical work on heterogeneity overlaps considerably with methodological research in ecology (biodiversity), economics (wealth inequality), statistics (namely measures of agreement), and statistical physics (generalized entropies). My primary clinical interests are mood disorders and neuropsychiatric sequelae of neurological and other somatic pathologies.

Research

Computational and Cognitive Neuroscience

• Computational theories of bipolar disorder physiology [forthcoming]
• Reinforcement learning in psychiatric populations (Nunes et al. 2018b)
• Learning reinforcement learning models from behavioural data (Nunes et al. 2018a)
• Goal-directed control (Fard et al. 2017)

Heterogeneity

• Measuring heterogeneity (Nunes et al. 2020, 2020a, 2020b, 2020f)
• Manifestations of heterogeneity relevant to psychiatry (Nunes et al. 2020a, 2020c, 2020d, 2020f)
• Heterogeneity in multi-centre machine learning studies (Nunes et al. 2018, 2020d, 2020e)
• Phenotypic heterogeneity of bipolar disorder and relation to treatment response (Nunes et al. 2020d, 2020e)

Machine Learning and Applied Computational Statistics

• Prediction of lithium response with clinical (Nunes et al. 2020e) and genomic data (Stone et al. [Submitted], Nunes et al. 2020d)
• Characterizing an invariant phenotype of lithium response (Nunes et al. 2020g, Nunes et al. 2020d)
• Neuroimaging based classification of bipolar disorder with machine learning (Nunes et al. 2018, Thompson et al. 2020)

Other Interests and Tangential Curiosities

• Technical: Computational verification of mathematical proofs; Fat-tailed distributions; Information geometry; The boundary between continuous and discrete spaces
• Heterogeneity related: Risks of "evidence-based" policies applied to heterogeneous populations
• Clinical psychiatry: Evidence-based psychiatric assessment; Clinical assessment of character strengths and virtues, and case formulation thereof