Computational Biology & Machine Learning
👋 Welcome! I’m Mahbuba Tasmin, a Ph.D. candidate in Computer Science at the University of Massachusetts Amherst, advised by Prof. Anna Green. I work in the SAGE Lab (Statistical and Genomic Evidence Lab), where my research bridges machine learning, computational biology, and antibiotic resistance genomics.
My current focus is on building interpretable, biology-grounded models for predicting drug resistance in Mycobacterium tuberculosis — combining sequence-based deep learning, evolutionary augmentation, and causal variant discovery.
I’m particularly interested in:
- Genomic and protein-based ML for resistance prediction
- Causal interpretability and structural biology
- Cross-species learning and data augmentation
- Benchmark dataset design for biological ML models
🔬 Research Highlights
- BIG-TB Benchmark (17K isolates, 11 drugs): Developing a unified dataset and evaluation framework for resistance prediction across modalities.
- Resistance Forecast Project: Integrating structural, evolutionary, and machine-learning features to predict variant impact.
- Evolutionary Augmentation: Leveraging multi-species homologs to enhance sparse training data for protein-level models.
You can read more about these in my publications and projects.
📚 Teaching & Mentorship
I serve as a Teaching Assistant for CS520 (Software Testing) at UMass, where I help students design and evaluate test coverage, mutation analysis, and automated testing frameworks in Java. I also mentor undergraduate and master’s students in research on ML for biological sequences.
🧠 Beyond Research
Outside the lab, I enjoy clay crafts, photography, and event organization — from handmade air-dry clay bowls to campus community events. I also contribute to graduate student initiatives at UMass through organizing academic and social programs.
📄 Quick Links
This site is built with the Academic Pages template, powered by Jekyll and hosted freely on GitHub Pages.