Antibiotic Adjuvant Design for RecA
A project on using predictive and generative models to search for RecA inhibitors that could make antibiotics work better.
Open research pageSite notes~6.8K tokens
Nikhil Maturi
I work on AI for scientific problems, mostly in biology and medicine. I like models that help us ask better questions, not just models that score higher on a benchmark.
Featured project
Antibiotic Adjuvant Design for RecA
HELIX is a predictive and generative platform for discovering highly selective RecA inhibitors.
Research
Projects across molecules, imaging, cells, and clinical prediction
Generative Biology / Featured
Antibiotic Adjuvant Design for RecA
HELIX is a predictive and generative platform for discovering highly selective RecA inhibitors.
Scientific ML / ResearchNeural ODEs to Learn Bloch Dynamics and Replace Differentiable Simulators
Learning continuous-time imaging physics for inverse MRI reconstruction.
Computational Biology / ResearchComputational Investigation of hIAPP Oligomer Extracellular Cytotoxicity Mechanisms
A computational look at how hIAPP oligomers may damage beta cells.
Health AI / ResearchPortable Real-Time Machine Learning Asthma Management System for Exacerbation Prediction
A real-time prediction system for respiratory risk and asthma management.
Computer Vision / ResearchUltrafine DNA Bridge Detection in Rat Kangaroo Cells
Computer vision for faint DNA bridges in microscopy images.
Research Observatory
A living map of AI for science work
An interactive page-section that treats research as a connected field: molecules, imaging physics, cellular mechanisms, and health systems as a measured instrument. Click or tap the satellites to switch projects.
InteractiveClick the satellites
AIfor science
Writing
Notes on models, biology, and ideas I’m still working through
Essays are Markdown-backed. Some include small interactive figures when a static diagram would hide the interesting part.
2026-05-04 / 8 min
Biological Graphs Need World Models
Why graph neural networks feel right for biology, why they still underdeliver, and what a better AI-for-biology architecture might preserve.
2025-05-11 / 22 minInside Agentic Tool Use
A visual explainer of what happens between an LLM API request, tool execution, repeated inference, and the final streamed response.
2025-05-06 / 18 minAI Coding Assistant Notes
Notes from using AI coding tools: mindset, prompting tactics, verification, context management, and working in real codebases.