Microbial electrochemical systems · research intelligence
MESSAI unifies a curated microbial-electrochemistry corpus, a causal parameter model, and a calibrated ML stack — so you can discover what’s known, design a reactor in 3D, and predict performance with honest uncertainty.
Live 3D · Drag to rotate · dual-chamber microbial electrolysis cell
What researchers can do here
Every surface reads from the same curated corpus and the same calibrated model stack — so what you discover, analyze, and predict all trace back to source.
Drive an interactive 3D microbial fuel cell and open a sourced dossier on every layer — its material, role, and the literature-traced fix a rigorous review caught — then compare it against reference designs.
See the case study →Browse canonical parameters with per-system literature distributions, hierarchical priors, and confidence — not single-number folklore.
Open the catalog →See how parameters depend on one another through a force-directed knowledge graph — fitted causal edges, correlations, and provenance in one interactive view.
Explore the graph →Build a bioelectrochemical reactor in the 3D lab and get live performance predictions with calibrated uncertainty and mechanistic loss breakdowns.
Open the lab →Fix your operating conditions, sweep one variable, and watch predictions move against the empirical distribution from the corpus.
Run a sweep →A research-scan view of physics-violating results, statistical anomalies, and Bayesian outliers worth a closer look.
Open the hunter →How it works
The same pipeline feeds every surface. Each stage is auditable — you can trace any number on the platform back through calibration, canonicalization, and extraction to the paper it came from.
19,000+ MES papers in the corpus, with full-text PDFs content-addressed and embedded for semantic retrieval.
Two-tier LLM extraction pulls measurements with their conditions. Every value keeps its DOI, and about half keep a source snippet.
Raw values are normalized to SI and mapped to canonical parameters, so numbers from different papers become comparable.
Hierarchical Bayesian priors (PyMC, log-normal) per system class, wrapped in conformal intervals with measured coverage.
Per-physics-family predictors return every output as {value, unit, ci_low, ci_high, confidence, source}. Interval fields are filled where a calibrated interval exists.
Design a reactor and see the prediction, the empirical evidence behind it, and the mechanistic loss decomposition side by side.
Calibrated prediction
Literature measurements are normalised to SI, pooled into a hierarchical log-normal prior per system class, and wrapped in conformal intervals with measured coverage where calibration exists. You get an interval, not folklore.
Causal structure
Parameters are linked in a directed acyclic graph. Each edge is labelled by how it was established — derived from physics, or fitted from co-observations in the corpus — so you can tell a mechanism from a correlation.
Open by default
The data and tooling that power MESSAI are released as standalone, citable packages — each with a schema, a license, and a CITATION.cff. Use them in your own analysis; they don’t depend on the platform.
The canonical MES parameter ontology — variables with literature distributions, a causal DAG, and full provenance.
CC BY 4.0DFT-computed electrode, membrane, and catalyst properties with Materials Project provenance and Pourbaix stability.
CC BY 4.0Curated MES-relevant microorganisms — electrogens, electrotrophs, community partners, and negative controls.
MITMethodology tooling and a physics-constraint validator for experimental design, sample sizing, and QC (Python).
MITPhysics-based models — Butler-Volmer kinetics, biofilm growth, and Nernst / Monod calculators.
MITEducational calculators — sustainability, life-cycle assessment, techno-economics, and SDG alignment.
CC BY 4.0A multi-agent research orchestration framework for literature analysis and insight generation.
MITResearch-gap identification and hypothesis generation with confidence scoring and contradiction detection.
MITA slug-keyed catalog of curated open MES datasets (Zenodo, Figshare), cross-referenced to Parameters and Materials.
CC BY 4.0New to the packages? Start with the manual or read how the parameter package is consumed downstream.
Two front doors · one platform
If you operate a waste stream
Pick a waste stream → see predicted recovery
Six industrial archetypes + custom influent characterisation. Live ML predictions with 95% CI.
Honest TEA + LCA per archetype
CapEx, OpEx, LCOE, LCO-water scaled with flow. GWP + fossil energy from peer-reviewed baselines.
5-physics-family routing
MFC / MEC / MDC / MES / MNRC / MMRC / MBES auto-routed by influent characteristics.
If you research MES
19,000+ papers · 201,954 measurements
Two-tier extraction (paper header + per-measurement conditions). Every row carries its DOI; about half carry a source snippet.
835 canonical parameters, FK-linked DAG
2,819 mechanistic edges with literature support. Multi-axis taxonomy (17 primary types × subtypes × combinations).
Per-class predictors + conformal bounds
97.98% OOS coverage at the 95% target on 940 held-out observations. Hierarchical priors fit via PyMC NUTS.
Get started
Search the corpus, design a reactor, and get predictions that carry their own uncertainty and cite their own sources.