Machine learning,
written into the genome.
I’m Abid Hasan — a bioinformatics and machine-learning scientist at Roche. I design deep-learning systems that turn raw sequencing data into clinical insight, from variant calling and MRD detection to large language models in healthcare.
I work where computational biology meets modern AI.
My work focuses on building accurate, efficient methods for analyzing high-throughput sequencing data — combining classical statistical learning with deep learning to solve real problems in genomics and molecular biology. At Roche I develop secondary-analysis pipelines for next-generation sequencing, and I’m endlessly curious about how the latest AI tools and LLMs can sharpen scientific discovery.
Recent work
A few projects from my research at Roche. The full archive — including publications and patents — lives on the research page.
Capabilities
The intersection of genomics, machine learning and software engineering — translated into working systems.
Deep Learning for Genomics
CNNs, autoencoders and gradient-boosted models for variant calling, MRD detection, imputation and classification.
NGS Secondary Analysis
End-to-end pipelines for next-generation sequencing — clustering, consensus, deduplication and quantification.
Algorithm Development
Performance-critical algorithms in C++, Java and Python for emerging sequencing chemistries and large datasets.
Applied AI & LLMs
Fine-tuning, RAG systems and LLM-assisted tooling that bring modern AI into research and clinical workflows.