ecoPrimals for Students, Lab Technicians, and Core Facilities

From: ecoPrimal — human + synthetic intelligence
Organization: 🔧🦎 ecoPrimals Date: March 17, 2026 Repositories: github.com/ecoPrimals — all AGPL-3.0-or-later

What This Is (30-Second Version)

🔧🦎 ecoPrimals is a collection of Rust programs that do the same things as Galaxy, QIIME2, NONMEM, and MassHunter — but faster, reproducible, and free. You run them on your own hardware. No cloud accounts, no Python environments, no Docker, no license keys.

If you have a laptop with a GPU (any gaming card works), you can run GPU-accelerated science. If you don’t have a GPU, everything runs on CPU too.

Getting Started (5 Minutes)

Step 1: Install Rust

curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh
# Follow prompts. Takes ~2 minutes.
# Restart your terminal, then:
rustc --version  # Should show 1.87+

Step 2: Clone the Spring You Need

Step 3: Build and Test

cd wetSpring/barracuda   # or healthSpring/
cargo test --workspace   # Runs ALL tests. Should see 0 failures.
cargo build --release    # Builds all binaries (release mode = fast)

Step 4: Run a Validation

# Pick one that matches your domain:
cargo run --release --bin validate_diversity         # Shannon, Simpson, Pielou, Chao1
cargo run --release --bin validate_dada2_full        # DADA2 denoising pipeline
cargo run --release --bin validate_pfas_decision_tree # PFAS screening
cargo run --release --bin exp001_hill                # Hill dose-response (healthSpring)
cargo run --release --bin exp075_nlme                # NONMEM/WinNonlin replacement

Every binary prints PASS or FAIL with explicit numerical checks. If everything passes, the pipeline is working correctly on your hardware.

For Genome Core / Sequencing Facility Staff

What This Replaces

Your current pipeline probably looks like:

Illumina sequencer → FASTQ → Galaxy/QIIME2 (Python/conda) → OTU/ASV tables → R (phyloseq/vegan) → diversity stats

🔧🦎 ecoPrimals collapses this to:

FASTQ → cargo run --release --bin validate_<experiment>

The Full 16S Pipeline (wetSpring)

Total: 306 validation binaries, 5,707+ numerical checks, 63 papers reproduced.

Why This Matters for a Core Facility

1. Reproducibility: Same binary, same result, every time. No Python version drift. No R package conflicts. No “it worked on my machine.”
2. Speed: GPU-accelerated spectral matching is 1,077× faster than CPU. Diversity calculations are GPU-native.
3. Provenance: Optional cryptographic signing on every result ( 🐻🐕 BearDog Ed25519). Maps to ISO 17025 traceability requirements.
4. No sysadmin: One cargo build compiles everything. No Galaxy server to maintain. No conda environments to debug.

For Graduate Students

What You Get

A real science pipeline validated against published papers — not a toy project. When you extend it with your own data, the infrastructure guarantees correctness.

The K-Nome Approach

K-Nome (Knowledge-Numeric Observed & Mentored Evolutionary Programming) is how this was built: one human with domain expertise mentoring AI (Cursor IDE) through iterative cycles. The Rust compiler is the fitness function — code either compiles and passes tests, or it doesn’t.

What this means for you:

• You don’t need to be a Rust expert. The compiler teaches you.
• Every module has tests that serve as executable documentation.
• The validation binaries are the ground truth — if your modification breaks a check, you know immediately.

Student Project Ideas (Real Science, Not Homework)

Each of these is publishable. The spring’s existing validation infrastructure guarantees your results are correct.

For Lab Technicians / Research Associates

What You Need to Know

You don’t need to write Rust. The validation binaries are pre-built executables. Your workflow is:

1. Prepare your data (FASTQ, mzML, CSV — whatever your instrument produces)
2. Run the relevant binary (cargo run --release --bin validate_*)
3. Check the output (PASS/FAIL with numerical tolerances)

If something fails, the output tells you exactly which check failed and what the expected vs actual values were.

Common Lab Data Formats Supported

Troubleshooting

Hardware Requirements

The entire ecosystem was built on ~$15,000 of consumer hardware. No HPC required for development or validation. ICER access expands scale, not capability.