Clinicians — Your Data Is Ready. Come Get It.

This page exists because someone needs to tell you.

There is a bioinformatics stack sitting on GitHub right now — fully open, fully validated, fully free — that does things your $50,000/year software licenses can’t do. It runs on your laptop. It doesn’t phone home. Your patient data never leaves your machine.

Nobody told you because nobody’s selling it. There’s no sales team. No conference booth. No rep buying you lunch. Just a microbiologist who built it and is telling you now: come get it.

What’s Ready

🧬 Protein Structure Prediction — helixVision

What it does: AlphaFold2/3 — the entire architecture — in pure Rust with f64 precision.

All 6 AlphaFold primitives implemented and validated:

• Evoformer (MSA processing)
• IPA (invariant point attention — the geometric reasoner)
• Diffusion (structure generation)
• Pairformer (pair representation processing)
• Confidence (pLDDT, PAE quality scores)
• Recycling (iterative refinement)

Pipeline: FASTA → structure → confidence → provenance. Every prediction gets a cryptographic provenance chain — you can prove which sequence produced which structure, when, on what hardware.

Target benchmark: 8.3 million LTEE structure predictions. ~$1,000 on 4× RTX 4070 vs ~$83,000 on cloud. Same physics. 83× cheaper.

git clone https://github.com/syntheticChemistry/neuralSpring
cd neuralSpring
cargo run --release --bin validate_alphafold_primitives
# 154 checks pass at 1e-10 tolerance

💊 Sovereign Pharmacometrics — healthSpring

What it does: Full PK/PD modeling. Population pharmacokinetics. Drug discovery screening. Replaces NONMEM ($3,600/year) and Monolix ($5,000/year).

For pharmacologists: Your NONMEM license costs $3,600/year. Your Monolix license costs $5,000/year. Both require specific operating systems, specific compilers, and specific training. healthSpring does the same math in Rust. It compiles on any platform. It costs $0. And your patient data never leaves your machine.

git clone https://github.com/syntheticChemistry/healthSpring
cd healthSpring
cargo run --release --bin validate_pkpd

Interactive Explorer: The Gonzales Interactive Explorer gives you IC50 curves, PK decay, tissue geometry, and hormesis — live charts, all computed locally.

🦠 Sovereign Metagenomics — wetSpring

What it does: Complete 16S rRNA pipeline. FASTQ to taxonomy to diversity to UniFrac. All in Rust. Replaces Galaxy and QIIME2.

7,083+ validation checks. Validated against 170 NCBI metagenomes and 11.9M paired-end reads. Python→Rust parity at machine-epsilon precision.

For microbiologists: You’re currently waiting in a QIIME2 Conda environment that takes 45 minutes to install and breaks every Python update. Or you’re uploading your patient sequences to Galaxy (someone else’s server). wetSpring compiles in 3 minutes and runs locally. Your sequences. Your machine. Your results.

git clone https://github.com/syntheticChemistry/wetSpring
cd wetSpring
cargo run --release --bin validate_16s_pipeline
# 7,083 checks — FASTQ → taxonomy → diversity → UniFrac

Also in wetSpring: LC-MS feature extraction and PFAS screening. If you’re doing analytical chemistry, the mzML parser and peak detection are validated and ready.

🧪 Drug Repurposing & Immunology

Anderson in Immunological Signaling — a published baseCamp paper using the Anderson localization framework applied to immune signaling networks. The same math that predicts electron transport in disordered materials predicts cytokine cascade behavior.

What clinicians get:

• Drug repurposing pipeline — screen existing approved drugs against novel targets using physics-based scoring
• Atopic dermatitis modeling (Anderson-immune framework)
• Cytokine signaling network analysis
• IC50 curve fitting with confidence intervals

This isn’t molecular docking. It’s physics. The same Anderson model that localizes electrons in disordered crystals localizes signals in disordered immune networks. It was validated on known drug-target pairs first.

🫀 Biosignal Processing

For cardiologists and neurologists:

All validated. All local. All Rust. No cloud. No PHI exposure. No HIPAA risk from third-party processing.

🌱 Precision Agriculture — airSpring

For agricultural scientists and agronomists:

3,123+ checks. 57 papers reproduced. 2,777+ individual validation points.

⚛️ Physics — For When You Need the Math

hotSpring — if you need molecular dynamics, nuclear equations of state, or lattice QCD:

700+ checks. 10+ papers reproduced. The physics is the same — the hardware is $4,000 instead of $4,000,000.

The Full Stack — By the Numbers

How to Start (Right Now, Tonight)

If you’re a clinician with patient data:

# Protein structure prediction
git clone https://github.com/syntheticChemistry/neuralSpring
cargo run --release --bin validate_alphafold_primitives

# PK/PD modeling (replaces NONMEM)
git clone https://github.com/syntheticChemistry/healthSpring
cargo run --release --bin validate_pkpd

# Your data stays on YOUR machine. Always.

If you’re a microbiologist with sequences:

# Sovereign 16S pipeline (replaces QIIME2/Galaxy)
git clone https://github.com/syntheticChemistry/wetSpring
cargo run --release --bin validate_16s_pipeline

# LC-MS and PFAS screening
cargo run --release --bin validate_lcms

If you’re a pharmacologist:

# Drug repurposing pipeline
git clone https://github.com/syntheticChemistry/healthSpring
cargo run --release --bin validate_drug_discovery

# Interactive IC50 / PK decay explorer
cargo run --release --bin gonzales_explorer

If you’re a data scientist with environmental samples:

# 100 Michigan weather stations, 80 years
git clone https://github.com/syntheticChemistry/airSpring
cargo run --release --bin validate_et0

Requirements:

• Rust (install: curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh)
• A Vulkan-capable GPU (any NVIDIA/AMD/Intel from the last 5 years)
• That’s it. No CUDA. No Python. No Conda. No Docker. No cloud account.

Why This Exists

I’m a microbiologist. Wayne State University. I studied cell membranes and metagenomics.

Then I needed to analyze data and discovered that every tool was either:

• Proprietary ($3,600/year for NONMEM, $5,000/year for Monolix, $50,000+ for enterprise genomics)
• Cloud-dependent (your patient data goes to someone else’s server)
• Fragile (Python environment breaks every update, Conda takes 45 minutes to install)
• Vendor-locked (CUDA only — NVIDIA takes 50% margin on scientific GPUs because they can)

So I built the alternative. From scratch. In Rust. With 20,695 validation checks to prove it works.

Every tool that was free, I used. Every tool that was locked, I replaced. Every claim, I validated against published papers. Every result, I made reproducible with cargo run.

3.6 million lines of Rust. One person. No funding. No institution. No grant. Just the code and the science and the absolute refusal to pay rent on tools that should be free.

Who Built This

Kevin Mok. Microbiologist. Not a CS grad.

Studied cell membranes and applied the architecture to software. The selective permeability of a lipid bilayer became the selective permeability of a firewall. The epitope recognition of an antibody became the visitor classification of a membrane. The immune response of a body became the threat detection of a system.

3.8 billion years of evolution is a better teacher than 4 years of computer science.

But the code is validated. 92,454 test functions. 175+ papers reproduced. Every check has a reference paper. Every result has an expected value. Every pipeline has a bit-exact parity test against the established tool it replaces.

You don’t have to trust me. Trust the checks. Run them yourself.

The Invitation

If you’re a clinician, a researcher, a pharmacologist, a microbiologist, an agronomist, or anyone who works with scientific data and is tired of paying rent on proprietary tools that own your results:

Your data is ready. Come get it.

• Code: github.com/syntheticChemistry (all springs)
• Infrastructure: github.com/ecoPrimals (all primals)
• Live site: primals.eco
• Contact: eco.primal@pm.me
• License: AGPL-3.0 — free forever, share-alike

No sales call. No demo request. No trial period. No credit card.

git clone and cargo run. That’s the onboarding.

Verify Everything

Every claim on this page links to runnable code:

Don’t believe me. Verify me. That’s the point.

The data is ready. The tools are free. The validation is done. The only thing missing is you.