From Concept to Reality: Building a Quantum-Infused Drug Discovery Engine

I’m under no illusion that building a working drug discovery engine is easy — even with the rise of AI. It’s a monumental challenge that demands mastery of multiple disciplines, months (if not years) of development, and relentless testing.

Even if you have a background in software engineering and have built complex systems before, creating a quantum-infused, production-grade drug discovery platform is a different league entirely. It means understanding not just software and cloud infrastructure, but also computational chemistry, quantum computing hardware, entropy management, and a laundry list of other specialties.

What You Need to Know to Even Attempt This

To design an engine like this, you’d ideally have:

• Above-average software engineering skills (Python, Node.js, containers, orchestration tools like Kubernetes)

• Solid knowledge of cloud runners (Cloud Run, ECS, etc.) and infrastructure

• Familiarity with quantum mechanics and quantum computing platforms (D-Wave, IonQ, QRNGs)

• A working grasp of computational chemistry concepts

• Enough experience with scientific APIs, containerized workloads, and distributed computing to glue all these pieces together

No, you don’t need to be a PhD chemist or physicist. But you do need to know how to collaborate with experts and extract what you need — from protein preparation protocols to docking score interpretation — to feed into your engine.

Where I Started

After 100 days of intense Node.js and browser-side development, I had the framework in place:

• A quantum entropy pipeline already running on Google Cloud Storage (GCS)

• Nearly five years of self-driven quantum computing R&D behind me — all without institutional backing

• A strong motivation: at 70 years old, I want to leave behind something useful, something accessible to anyone on Earth, that can help identify life-saving medicines

The vision is bold: produce a shortlist of potential drug candidates — from millions of compounds — and hand them to wet labs for real-world validation.

The Engine’s Purpose

To take in computational chemistry data, process it through classical and quantum algorithms, encode the results, and output a ranked list of promising compounds.

That means:

1. Designing a high-performance simulation system

2. Defining constraints and variables for each run

3. Running both classical and quantum-seeded algorithms

4. Outputting AI-ready datasets for analysis and filtering

I’ve done complex simulations before — in oil and gas — but drug discovery adds layers of scientific complexity.

Version 4: A Real, Working System

After three earlier prototypes, I now have Version 4 — the first to run true molecular docking on real proteins and ligands, with quantum entropy injection.

What’s in it?

✅ Real Molecular Docking – AutoDock Vina, a gold-standard docking algorithm used in academic and industry research

✅ Real Protein Preparation – Using PDBFixer (fix missing atoms/residues), PDB2PQR (electrostatics), PropKa (pKa/protonation), and AutoDockTools (ADT) for receptor file generation

✅ Real Chemical Data – PubChem ligands and RCSB PDB protein structures — the same datasets pharma companies use

✅ Real Quantum Entropy – From the Australian National University QRNG and my own USB/PCIe QRNG devices, with proprietary extraction logic, all stored in my GCS bucket and soon to be directly connected to the D-Wave QPU

✅ Production-Grade Infrastructure – Cloud Run containers, GCS storage,

Supabase tracking, and manifest-driven data management

Why This is Scientifically Valid

The toolchain itself is the same one used in academic papers. The difference between my setup and a $100K+ commercial license is speed and some advanced features — not the core science.

The free stack (Vina + open-source prep) vs. the licensed stack (Schrödinger, OpenEye) is like Linux vs. Windows: different economics, same fundamentals.

What’s Next for Live Production

• Deploying containers at scale – Dedicated docking workers pulling tasks from a queue, running in Cloud Run or Kubernetes

• GCS bucket organization – Separate paths for protein assets, ligand shards, docking results, and analytics parquet files

• Secrets & config – VINA_CLOUD_RUN_URL, entropy bucket paths, API keys for licensed services when available

• Crowd-powered runs – Volunteer compute nodes processing small shards (desktop, laptop, even mobile) that add up to a massive distributed effort

Why Quantum Matters Here

Traditional simulations seed their random number generators with deterministic PRNG outputs. I seed mine with true quantum entropy, introducing a fundamentally different collapse landscape.

In tornado forecasting, this quantum approach has shown directional biases in collapse zones vs. PRNG — and I’m now seeing similar patterns in drug docking. In some cases, QRNG-seeded runs appear to favor binding conformations that PRNG runs miss entirely.

Entropy comes from my QRNG hardware — USB and PCIe devices — controlled with Python scripts, harvested into GCS, and injected into simulations.

Bottom Line

This is no longer just an “interesting simulation.” It’s a real drug discovery engine, capable of producing publishable results, and built so it can scale — with the public’s help — into a global, quantum-powered search for life-saving medicines.

The R&D continues.

So, if you like to collaborate, hey! send me an email. if you are an investor, well, send me an email: videomover@gmail.com.

I do have a Phone, but I never answer it! - Just send me an email.