I Built a Real Quantum-Infused Drug Discovery Platform — Here’s Where I Am Now
- mansour ansari
- Aug 12
- 3 min read
After months of tense work, countless late nights interacting with my “farm” of computers, and real quantum hardware (USB and PCIe QRNG devices), I’ve built something rare:
a fully functional, quantum-assisted drug discovery infrastructure.
We’re talking:
AutoDock Vina containers
Automated protein preparation orchestration
GCS (Google Cloud Storage) workflows
Cloud-run ready pipelines
This isn’t a “simulation of a simulation.” This is production-grade computational chemistry infrastructure.But I’ve hit a cost wall — and that’s why I’m opening the door to collaboration.
The Licensing Roadblock
High-end computational chemistry APIs from OpenEye and Schrödinger are incredibly powerful — but also incredibly expensive. I don’t have the budget for their licensing fees yet. I am a lone quantum researcher/developer in Oklahoma making an attempt to build a system that can hopefully - some day - very soon - help identify new cancer drugs.
Instead of stalling, I pivoted to what I call my no-license MVP: a completely free, scientifically validated toolchain that any academic lab would recognize.
It’s modular — meaning licensed tools can be slotted in later with no re-architecture.
The Core Free & Validated Stack
✅ Free Protein Prep ChainRCSB PDB → pdbfixer → PDB2PQR/PropKa → ADT → receptor.pdbqt
RCSB PDB – The world’s open-access archive of protein 3D structures.
pdbfixer – Fixes missing atoms/residues, removes unwanted parts, prepares for simulation.
PDB2PQR – Assigns partial charges & atomic radii for electrostatics.
PropKa – Predicts pKa and sets correct protonation states.
ADT (AutoDock Tools) – Converts to Vina-ready PDBQT with grid definitions.
✅ Free Ligand PrepRDKit → meeko → ETKDG conformers → QED/PAINS filtering
RDKit – Cleans, normalizes, deduplicates molecules from PubChem or other libraries.
meeko – Converts ligands to docking-ready PDBQT with torsional data.
ETKDG – Generates realistic 3D conformers for better docking accuracy.
QED/PAINS filtering – Removes false-positive-prone or poor drug-like candidates.
✅ Enhanced DockingVina + QuickVina2 + Gnina (AI rescoring)
Vina – The gold-standard docking engine.
QuickVina2 – A faster fork for large-scale screening.
Gnina – AI rescoring using deep learning for more accurate pose ranking.
✅ Free Molecular Dynamics (MD)OpenMM minimization / short relax
Relaxes docked poses to remove clashes and improve physical realism.
✅ Free ADME/Tox FilteringADMET-AI or RDKit + XGBoost
Predicts drug-likeness, toxicity risks, and other pharmacokinetic red flags.
Why This Stack Matters
Actually functional — can produce real, reproducible docking results.
Scientifically defensible — same core tools used in peer-reviewed research.
Upgradeable — licensed tools can drop in later for speed boosts and advanced features.
Investor-friendly — operational now, scalable later.
Uniquely differentiated — quantum entropy injection + citizen-science compute layer.
The Quantum Twist
Most drug docking pipelines use PRNG (pseudo-random number generators) for simulation seeding.I use true quantum entropy from my QRNG hardware, harvested and injected at docking and MD stages.
In my tornado forecasting work, I’ve seen that QRNG seeding produces different collapse patterns compared to PRNG — sometimes with a directional bias toward better solutions.I’m applying that same principle here, even to FDA-approved drug docking replays.
What I Need Next
This MVP proves the concept. Now, to push it into full-scale, investor-ready production, I need:
Licenses for OpenEye and Schrödinger APIs (protein prep, ligand generation, high-speed docking).
GPU/CPU cluster funding for high-throughput docking and molecular dynamics.
Strategic collaborators with domain expertise in computational chemistry, structural biology, and AI drug design.
Investors or partners willing to accelerate deployment in exchange for equity or shared IP rights.
Call for Collaboration
If you are:
An academic with HPC access and interest in quantum-assisted docking,
A pharma/biotech professional looking for innovative lead generation methods,
An investor who sees the potential in marrying quantum entropy with AI-driven drug discovery,
…I want to hear from you.
This is not just a concept. It’s running today — and with the right partners, it can discover real drug candidates, at scale.
📧 Contact me directly to discuss technical details, investment opportunities, or collaboration models. videomover@gmail.com
Comments