Theoretical Physicist · Computational Biologist · Scientific Software Architect
Molecular Simulation • Statistical Mechanics • Scientific Computing • AI for Molecular Science

Co-directing computational biophysics research and architecting robust scientific software platforms: from non-equilibrium statistical mechanics to accelerated compute kernels, scientific Python APIs, and TypeScript visualization.

I work at the intersection of theoretical physics, computational molecular science, and software platform architecture. I formulate physical models of biomolecular kinetics and construct the modular software abstractions required to translate mathematical theory into reproducible, high-throughput pipelines.

I co-direct the Computational Biology and Drug Design Research Unit (UIBCDF) and lead the architecture of MolSysSuite, an open-source scientific computing platform for molecular modeling, simulation, and AI-assisted workflows.

Core Technical Foundation:

• Languages & Production Stack: Libraries and platform built across Python (scientific stack, data science, biophysical modeling), Rust (memory-safe accelerated compute kernels), and TypeScript (modern web visualizers and frontends). Formative background in native Fortran for classical numerical physics.
• Engineering & Delivery: Automated CI/CD pipelines, agent-ready testing tooling, Linux environments, Conda packaging, and multi-platform distribution.

Rather than a loose collection of research scripts, MolSysSuite is engineered as an agent-ready, layered scientific computing architecture designed to reduce glue-code complexity and manage interoperability across heterogeneous molecular-science ecosystems:

• [ AI & Orchestration ] → molsys-ai
• [ Domain & Modeling ] → MolSysMT · TopoMT · MolSysViewer · ElasNetMT
• [ Core Contracts & Telemetry ] → PyUnitWizard · ArgDigest · DepDigest · SMonitor
• [ Native Compute & Automation ] → Rust Accelerated Kernels · TypeScript · pytest-receptor · Conda Packaging

• MolSysMT: Unified molecular systems toolkit providing syntactic consistency and transparent conversion across 10+ heterogeneous engines and formats (OpenMM, MDTraj, MDAnalysis, PDB, etc.), powered by accelerated compute backends in Rust.
• molsys-ai: Agentic interface translating high-level biophysical queries into structured, inspectable, and reproducible MolSysSuite execution pipelines.
• TopoMT: Geometric and topographic characterization of molecular surfaces, binding pockets, cavities, and transport tunnels.
• MolSysViewer: High-performance 3D molecular visualization widget for Jupyter environments built on Mol* with a clean, programmatic API in TypeScript and Python.

• PyUnitWizard: Universal adapter and orchestration layer for physical quantities and unit libraries (Pint, OpenMM Units, Unyt), ensuring dimensional safety across codebases.
• ArgDigest: Contract-based argument auditing, type normalization, and introspection layer to decouple input validation from scientific logic and enable deterministic agent introspection.
• DepDigest: Lazy-loading dependency manager eliminating startup overhead when interfacing with heavy optional scientific libraries.
• SMonitor: Centralized telemetry, structured diagnostics, and event routing layer across heterogeneous Python libraries.
• pytest-receptor: Specialized test reporter for CI/CD and coding agents—compact, root-cause-grouped verdicts optimized for reliable interpretation by coding agents.

• Peer-Reviewed Science: Author of 30+ publications in computational biophysics and statistical mechanics with 1,000+ total citations.
• Key Publications:
  • Conformational Markov Networks: Exploring the free energy landscape: from dynamics to networks and back (PLoS Comput. Biol., 2009) — Pioneering graph-theoretical framework mapping molecular trajectories into discrete kinetic networks.
  • Allosteric & Receptor Dynamics: Conformational transitions and activation mechanisms in GPCRs (Biophys. J.) — Kinetic decomposition of activation pathways and free energy landscape transitions.
• Architecture & Interoperability: Designed cross-engine bridges in MolSysMT connecting over 10 major structural biology formats without mandatory vendor lock-in.
• Ecosystem Governance: Maintainer of the UIBCDF Conda channel, orchestrating automated multi-platform builds, reproducible recipes, and environment deployment for open science.
• Open Source Stewardship: 10+ actively maintained repositories under @uibcdf distributed under permissive open-source licenses (MIT / LGPL).

• Unit Co-Direction: Co-PI and Senior Researcher at the UIBCDF (Hospital Infantil de México Federico Gómez, Mexican National Institutes of Health).
• People & Project Leadership: Supervised and mentored graduate researchers, postdocs, and technical staff across molecular simulation, statistical physics, and scientific software engineering.
• Research Program: Focused on mechanistic computational biology:
  • Statistical Mechanics & Molecular Kinetics: Conformational Markov Networks, transition networks, metastable states, and non-equilibrium free energy surfaces.
  • Computational Biophysics & Simulation: Multiscale dynamics of allosteric regulation, GPCR activation pathways, and membrane biophysics.
  • Molecular Modeling & Drug Discovery: Quantitative thermodynamic binding validation, cell-penetrating peptides (CPPs), and structure-guided pharmacological design.

• Physics-Aware by Design: Abstractions preserve physical units, structural semantics, conservation laws, and reproducible statistical sampling.
• Contract-Driven & Agent-Ready: Clean separation between validation contracts (ArgDigest), runtime performance, and business logic to ensure deterministic behavior for human developers and autonomous coding agents.
• Reproducible Packaging & Delivery: Automated multi-platform CI/CD testing, structured telemetry, and unified Conda distribution to eliminate setup barriers.

📍 Laboratory: UIBCDF — Hospital Infantil de México Federico Gómez (Mexico City)
✉️ Contact: diego.prada.gracia [at] gmail.com