Design Experiments From Reported Results

 Quick Analysis Plan Copied

 Long Analysis Plan

Evidence Basis and the Confound

The cited toxicology study tested whole powders — fungus grown on oat substrate — at 0, 500, 1,000, and 2,000 mg/kg bw/day in Wistar rats (10/sex/group, OECD TG 408), concluding NOAELs of 2000 mg/kg bw/day for both products 90-day NOAEL evidence for both powders. Because no oat-only arm was included, any contribution from oat substrate, residual mycelial digestion of oats, or substrate-modulated fungal metabolites is unresolvable from these data — a blindspot the authors themselves acknowledge (only oat-grown products were tested; no other substrates were in scope).

Proposed Disentangling Design

• Factorial 2×2 in vivo study: (a) organic oat powder alone, (b) same fungus grown on a defined liquid/solid non-oat substrate, (c) the commercial oat-cultured powder, (d) vehicle control — each at doses matched to 2000 mg/kg bw/day and dose-matched on a total-mass and a fungal-biomass-normalized basis.
• Substrate-matched controls: sterilized, spent (non-inoculated) oat substrate processed identically to the commercial powder isolates thermal, enzymatic, and microbial-degradation artifacts of processing.
• Chemical attribution: untargeted LC-MS/MS metabolomics + GC-MS on all four arms, with oat-marker metabolites (avenanthramides, β-glucan) versus fungal markers (erinacines/hericerins for H. erinaceus; PSK/Krestin precursors, polysaccharides for T. versicolor) quantified by targeted HPLC to compute a fungal:substrate contribution index per endpoint H. erinaceus bioactive compound chemistry review.
• Bioinformatic analysis: multivariate (PLS-DA, WGCNA in Python scanpy/sklearn) linking metabolite loads to OECD endpoints (hematology, clinical chemistry, organ weights, histopathology); statsmodels for dose-response modeling and NOAEL benchmark-dose (BMD) re-estimation per arm.
• Mechanistic triangulation: compare against H. erinaceus proteome-regulated terpenoid pathways (fruiting body vs mycelium differential expression) to predict which metabolites track substrate choice H. erinaceus proteome bioactive metabolite regulation.

Confidence and Blindspots

Confidence: moderate. The factorial design is the only rigorous way to partition toxicity attribution, but it is BGPT-inferred — no published oat-only arm exists in the cited work. Missing evidence: substrate composition data for the commercial powders, sex-specific acute data (only females tested acutely), and any chronic/carcinogenicity data. Sponsor funding by M2 Ingredients introduces potential sponsor bias that independent replication should address Sponsor funding and scope limitations. Falsification: if oat-only and fungus-only arms reproduce the full NOAEL profile independently, neither factor alone dominates and interactions dominate — testable with a two-way ANOVA/BMD hybrid model.