About

Nivedita Bhadra

Senior Computational Scientist working at the intersection of physics, statistical modeling, and biomedical data.

I trained as a physicist, then carried that quantitative foundation into biomedical research and statistical genetics — building models that have to hold up against real, messy, large-scale data.

Background

I’m currently at the Institute of Biological Psychiatry, Denmark, where I build statistical genetics models for risk prediction on large-scale cohort data — liability-scale modeling, polygenic risk scores, and related quantitative genetics methods.

Before that, I spent several years at the Translational Genomics Research Institute (TGen), applying high-dimensional statistical modeling to biomedical and clinical data.

My path into this work started in physics: an M.Sc. from IIT Delhi, followed by a PhD in Computational Physics from IISER Kolkata, where I built numerical models of complex physical systems. That grounding in simulation, statistical inference, and large-scale computation is the thread connecting everything since — whether the system is physical, genomic, or textual.

Education

PhD, Computational Physics Indian Institute of Science Education and Research (IISER), Kolkata, India — 2012–2018 Read a summary of the doctoral research →

M.Sc., Physics Indian Institute of Technology (IIT), Delhi, India — 2009–2011

M.Eng., Big Data Analytics Arcada University of Applied Sciences, Finland — 2024–2025 Read a summary of the Master’s research →

Publications

A full list of peer-reviewed papers and theses — spanning biomedical data science and computational physics — is available on the Publications page.

Technical Skills

Programming & Tools

Python R SQL Bash / awk Git & GitHub Quarto

Statistical & Quantitative Methods

GWAS Polygenic Risk Scores Liability-Scale Modeling REML SEM SVD Mendelian Randomization Colocalization Hypothesis Testing

Machine Learning & NLP

Model Evaluation Transformers Local LLMs Sentiment Analysis Topic Modeling

High-Performance & Reproducible Computing

HPC / Cluster Computing Snakemake Dask Reproducible Pipelines

Elsewhere