Available for positions

Proteins move.
Most tools don't.

Postdoctoral researcher at UNH. Proteins are never still. Neither am I — I chase their conformational states across systems, from photoreceptor enzymes to viral targets, using MD simulation and integrative modeling to catch what crystallography misses in motion.

9
Peer-reviewed publications
2
Book chapters
10+
Years MD simulation
5+
Disease targets studied
PDE6 GAFab Regulatory Domain PDB 6X88 · Cone photoreceptor · 3.2 Å
Molecular Dynamics· AlphaFold2· Integrative Modeling· Allostery· IMP · MODELLER· XL-MS· GROMACS· QSAR· Drug Discovery· Network Analysis· Agentic AI· PDE6· Phototransduction· Molecular Dynamics· AlphaFold2· Integrative Modeling· Allostery· IMP · MODELLER· XL-MS· GROMACS· QSAR· Drug Discovery· Network Analysis· Agentic AI· PDE6· Phototransduction·

The structure is solved.
Now what?

A crystal structure is a beautiful thing. It is also a photograph of a protein mid-blink — one moment, one conformation, one slice of a much richer story. The interesting stuff — the allostery, the conformational shifts, the rewiring of internal communication — happens in between.

That's where I live. I take experimental data — XL-MS, NMR, SAXS — and combine it with physics-based simulation to chase proteins across their conformational landscape. Then I ask what changed, why it changed, and whether anyone can do something useful with that.

Usually someone can. That's what keeps it fun.

🧬
AlphaFold2 + Homology
Seed structure from sequence
⚗️
XL-MS Restraints
Experimental distance constraints
🔧
IMP · MODELLER
Integrative refinement
All-atom MD · GROMACS
Conformational sampling
🕸️
Network Analysis
DCCM · GMI · Louvain communities
🎯
Druggable Allosteric Sites
Actionable structural insight

Four problems I've worked on

Different proteins, different diseases, same obsession — understanding what's happening when a protein changes shape.

PDB 6X88

PDE6 Holoenzyme — Four Conformational States

The same enzyme, four different stories. Using XL-MS restraints to guide integrative modeling with AlphaFold2, IMP, and MODELLER, I resolved how cone PDE6 looks in its apo, cGMP-bound, Py-inhibited, and fully-active states.

AlphaFold2IMPXL-MSGROMACS
Cone PDE6 · 6X88

Allosteric Networks & Druggable Nodes

Using DCCM, generalized mutual information, and community detection (Louvain/Newman), I map how the internal communication network of cone PDE6 reorganizes — and which nodes become druggable.

DCCMGMILouvainNetworkX
RPS6KB2
NRF2
RAC1
PI3K
Virtual Screening
MD Validation
Structure-Based

Structure-Based Drug Discovery — HNSCC

Four oncogenic targets in head and neck cancer. Virtual screening, docking, and all-atom MD to identify and validate small-molecule hits against RPS6KB2, NRF2, RAC1, and PI3K — ranked by binding free energy.

AutoDockGROMACSHNSCCDrug Discovery
EZH2 Inhibitors
HIV-1
Antimalarial
QSAR
XGBoost
SAR Networks
ML-Driven

ML-Driven QSAR — Oncology & Infectious Disease

Consensus QSAR models — RDKit fingerprints, XGBoost, random forests — to rank inhibitor activity across EZH2, HIV-1, and antimalarial targets, with network-based SAR clustering.

QSARRDKitXGBoostNetworkX
Conformational landscape
State I
Apo
State II
Ligand-bound
State III
Inhibited
State IV
Active

The central question

A system worth understanding

Over the years the proteins have changed — GPCRs, GTPases, kinases, viral enzymes — but the questions have stayed the same: how does a conformational change reshape the internal communication of a protein, and where does that create opportunity?

The same protein, four different states, four different stories. Mapping what changes between them — in structure, in dynamics, in the allosteric network — is where the interesting biology lives.

XL-MS · MD · Integrative Modeling · Network Analysis
9
Peer-reviewed publications
2
Book chapters
4
Conformational states mapped
10+
Years MD simulation
Structural Modeling
AlphaFold2 + Integrative (IMP/MODELLER)
Molecular Dynamics (GROMACS/AMBER)
XL-MS data integration
Network & Drug Discovery
Allosteric network analysis (DCCM/GMI)
ML-QSAR pipelines (RDKit/XGBoost)
Virtual screening & docking

Tools I use every day

Embedded in an experimental lab

Three years sitting next to biochemists has been a good education — in what questions actually matter, what data experimentalists trust, and how to make a simulation answer something real rather than just look impressive.

Building AI that reasons

Applying agentic AI patterns to scientific literature — autonomous agents that search, analyze, and summarize research.

Level 1 Project
ReAct Loop

PubMed Research Agent

A CLI-based agent that autonomously searches PubMed for any topic, retrieves publication counts by year, and summarizes key findings using Google Gemini — demonstrating tool use, autonomous reasoning, and the foundational ReAct pattern behind all modern AI agents.

Python 3.12 Biopython Google Gemini NCBI Entrez API Agentic AI
Searches
PubMed / NCBI
Summarizes with
Google Gemini
Pattern
ReAct Loop
View on GitHub →

More agentic AI projects coming — multi-tool research agents, autonomous coding agents, and multi-agent pipelines.

Selected work

Prep
The N-terminal motif of PDE6 promotes catalytic subunit dimerization and contributes to allosteric regulation by cGMP and the inhibitory gamma-subunit
Madhukar G., Wang X., Gupta R., Tyler M.E., Cote R.H. · In preparation
Prep
Molecular architecture of the rod photoreceptor RGS9-1/Gb5L-R9AP inactivation complex
Hagearty H.E., Madhukar G., Wang X., Cote R.H. · In preparation
Prep
Structural basis for the sequential activation mechanism of rod photoreceptor PDE6 by transducin
Wang X., Madhukar G., Cote R.H. · In preparation
2025
E3 ubiquitin ligases and their therapeutic potential in disease management
Madhukar G., Haque M.A., Khan S., Kim J.-J., Danishuddin · Biochemical Pharmacology, 236, 116875
2025
Network-based clustering and statistical evaluation to elucidate structure-activity relationships of EZH2 inhibitors
Danishuddin M., Haque M.A., Madhukar G., et al. · SAR QSAR Environmental Research, 36(9), 827-851
2025
Machine learning-driven consensus modeling for activity ranking and chemical landscape analysis of HIV-1 inhibitors
Danishuddin M., Haque M.A., Madhukar G., et al. · Pharmaceuticals (Basel), 18(5), 714
2024
Potential inhibitors of RPS6KB2 and NRF2 in Head and Neck Squamous Cell Carcinoma
Madhukar G., Subbarao N. · Journal of Biomolecular Structure & Dynamics, 42(4), 1875-1900
2022
Exploring GPR109A receptor interaction with hippuric acid using MD simulations and CD spectroscopy
Bhandari D., Kachhap S., Madhukar G., et al. · International Journal of Molecular Sciences, 23(23), 14778
2022
Identification of potent and novel inhibitors against RAC1: a Rho family GTPase
Madhukar G., Subbarao N. · In Silico Pharmacology, 10(1), 13
2022
In silico prediction of potential inhibitors against PI3K catalytic subunit alpha in HNSCC
Madhukar G., Subbarao N. · Journal of Biomolecular Structure & Dynamics, 40(10), 4697-4712
2021
Current and future therapeutic targets: a review on treating Head and Neck Squamous Cell Carcinoma
Madhukar G., Subbarao N. · Current Cancer Drug Targets, 21(5), 386-400
2024
Book Chapters
Unveiling the landscape of cancer: from therapeutic targets to their inhibition
Madhukar G., Subbarao N. · Recent Trends in Diabetes and Cancer Research and Its Management, 1(1), 72-131
2023
Current and potential pharmacological targets for therapeutic intervention in HNSCC
Madhukar G., Subbarao N. · Socio Scientific Interaction in Diabetes and Cancer and Its Management
View full list on Google Scholar

Let's work
on something worth solving

If your lab is chasing a hard biological question and computation is part of how you want to answer it — I'd genuinely like to hear about it. Positions, collaborations, projects that haven't found a home yet. All fair game.

Available immediately

What I'm
looking for

Seeking a lab tackling complex biological questions where computational strategy is a foundational partner from day one, rather than a post-script to the experiments.

  • Integrative structural biology labs
  • Computational drug discovery teams
  • Cryo-EM groups wanting to add dynamics
  • Viral protein structure & allostery
  • AI/ML applied to structural biology
Discuss a project →