P4 Batch 14 — Ingest Report (2026-05-13)

Positions 4568–4767 of the P4 manifest, sorted by (-year, handle). Year band: 2022. Two group-1 agents were launched (the first hit a usage-policy refusal after reading content from a clinical case report paper; the retry succeeded). One group-3 agent was also relaunched after a mis-triggered skill invocation.

Summary

GroupPositionsCommitPagesFPFP rate
g14568–4617e4c0b2334794%
g24618–4667011a29854590%
g34668–471767cbff764488%
g44718–4767364792f74386%
Total4568–47672117989.5%

The very high FP rate (89.5%) is consistent with recent batches in this year band. The 2022 P4 cohort contains a high density of molecular biology, materials science, sensor/optics, and plant physiology papers that cleared the triage manifest’s text-mining filter but lack food-matrix concentration data.

Source pages by group

G1 — 3 pages (positions 4568–4617)

  1. chavez-capilla2022-arsenolipids-seafood — CC BY review; arsenolipid species in seafood — tAs in seafood 5–100 mg/kg, arsenolipids in fish oils 0.2–16 mg/kg; As in breast milk of fish-eating mothers 0.3–4.46 µg/kg with 2–61% as AsFAs/AsHCs; directly strengthens the iAs/tAs non-substitutability argument for seafood and fish-oil product pages and corroborates the promoted organoarsenical-inertness-assumption synthesis page
  2. wade2022-cacao-cadmium-drivers — quantitative synthesis, 489 site-years, 2,127 observations, 10 countries; total soil Cd and soil pH are the dominant predictors of cacao bean Cd; available soil Cd adds no predictive power; EU 0.8 mg/kg limit frequently exceeded in South American origins; Côte d’Ivoire (world’s largest producer) absent from the literature — documented West Africa data gap
  3. sarkar2022-kathmandu-drinking-water — 35 Kathmandu Valley sources, AAS; As max 0.071 mg/L (7× WHO guideline), Mn max 3.229 mg/L (40× WHO health guideline), Al max 0.53 mg/L; consumer taste ratings uncorrelated with Mn/Fe, supporting the argument that aesthetic-based WHO Mn guidance is insufficient

G2 — 5 pages (positions 4618–4667)

  1. frisbie2022-arsenic-drinking-water-global — global review of As drinking water regulations across 176 countries; reference document for the regulations layer
  2. huhmann2022-arsenic-bangladesh-wells-rice — Bangladesh mass-balance study; n=814 rice samples, 50–1,200 µg/kg tAs; characterizes the rice As exposure pathway in a high-exposure population
  3. kumar2022-macrophyte-minerals-fish-feed — 12 freshwater macrophyte species analyzed by ICP-MS for Cd, Pb, Sn, Ni, Cr, Al; relevant to aquaculture feed ingredient contamination
  4. ravalli2022-uranium-metals-us-water-systems — 37,915 US community water systems; prevalence and co-occurrence of uranium and arsenic; uranium in US drinking water as a systematically underappreciated exposure
  5. efsa2022-pgpr-e476-heavy-metals — EFSA opinion on polyglycerol polyricinoleate (E 476 / PGPR); As, Pb, Hg, Cd specification limits for this common chocolate-thinning emulsifier

G3 — 6 pages (positions 4668–4717)

  1. efsa2022-polyglycerol-esters-e475-heavy-metals — EFSA opinion on E 475 (polyglycerol esters); 14 commercial samples, all Hg/Cd/As below LOQ, max Pb 0.29 mg/kg; Panel recommends lowering all four EU specification limits
  2. trueman2022-seasonal-lead-drinking-water-aluminum-halifax — seasonal Pb release in Halifax drinking water; aluminum from coagulation suppresses orthophosphate and raises Pb release 41% in coupon experiment (n=849 Al, n=193 Pb first-draw); mechanism-level evidence for coagulant chemistry as a Pb exposure driver in treated water
  3. gintamo2022-groundwater-quality-oromia-ethiopia — 17 boreholes, Oromia, Ethiopia; AAS for Pb, Cr, Fe, Cu, Mn, Zn; children HI 1.861 in dry season indicating non-carcinogenic risk from chronic groundwater consumption
  4. wang2022-heavy-metals-gastric-cancer-plasma-china — ICP-MS for 18 metals in plasma of 105 gastric cancer patients vs 62 controls, Changzhi, China; Pb, Cr, Hg, Sb, Sn, Tl elevated in cancer patients; Sb positively correlated with MSI status
  5. rarivoson2022-organic-resources-madagascar-trace-metals — n=1,000 Madagascar organic resource samples (manure, compost, agro-industrial waste); n=39 subset metal analysis; mean Pb 429 mg/kg DM, max 3,982 mg/kg DM; supply-chain contamination via organic soil amendments
  6. efsa2022-glycerol-e422-heavy-metals — EFSA opinion on E 422 (glycerol); 18 commercial samples all below LOD for all four toxic elements; Panel recommends lowering current EU specification limits

G4 — 7 pages (positions 4718–4767)

  1. knappett2022-arsenic-fluoride-aquifer-mexico — systems dynamics model projecting As in Guanajuato aquifer rising from 14 to 46 µg/L over 100 years under business-as-usual abstraction; long-term drinking water As risk in a major agricultural region
  2. bai2022-arsenic-coal-burning-shaanxi — coal combustion as food contamination mechanism; corn 4.24 mg/kg tAs, pepper 12.80 mg/kg — 21× and 256× Chinese GB2762-2005 limits; household coal burning in northwest China as a distinct As exposure pathway absent from standard food-chain surveillance
  3. li2022-mercury-isotope-biota-review — ESandT critical review; isotope-level mechanistic evidence that tHg vs MeHg speciation in marine biota is non-negotiable; >95% of tHg in piscivorous fish is MeHg; the mechanistic backbone for why tHg/MeHg speciation non-substitutability is defensible against analytical challenge
  4. bair2022-infant-toddler-food-heavy-metals-policy — Frontiers in Nutrition narrative review (B-tier); consolidates 2021 Congressional Subcommittee infant food findings; brands’ internal documents showed 200–400 ppb tAs in rice ingredients; FDA 100 ppb iAs action level for infant rice cereal; policy history reference for the infant food products page
  5. kubachka2019-arsenic-speciation-fruit-juice-fda — AOAC First Action Method 2016.04 (FDA HPLC-ICP-MS); the FDA reference method underlying the 10 ppb iAs action level for apple juice; critical methods reference for the testing pages
  6. oliveira2022-mercury-fish-pregnant-women-tapajos — Tapajos River basin, Brazil; piscivorous fish (tucunaré, pescada) exceed FDA/EPA 0.5 mg/kg tHg limit in at least one season; 48% of 110 pregnant women exceed US EPA hair Hg reference dose; Amazon artisanal fishing populations as a vulnerable group for MeHg
  7. agarwal2022-heavy-metals-vegetables-east-kolkata — sewage-fed wetland vegetables, Dhapa, Kolkata; ICP-MS for Pb, Cd, Cr, Hg; maize Pb 31.8–42.3 ppm DW; site-specific extreme contamination from sewage farming

New page proposals

No ingredient or product pages crossed the creation threshold from this batch. Notable near-thresholds:

Uranium in US drinking water — ravalli2022 (37,915 US community water systems) is the first uranium-specific study in the corpus focused on drinking water exposure rather than food matrices. If 2–3 more papers characterize uranium in food-chain pathways, a metals/uranium.md page could be justified.

Coal-burning as arsenic pathway — bai2022 (Shaanxi household coal burning) represents an exposure pathway distinct from standard agricultural contamination. No dedicated wiki page exists for this mechanism; it is currently covered in passing on metals/arsenic.md.

Synthesis threads

Mining-impacted freshwater fish (promoted candidate — 3 of 4 criteria met)

oliveira2022-mercury-fish-pregnant-women-tapajos adds Amazon basin data at the human-exposure level: 48% of 110 pregnant women in Tapajos communities exceeded the US EPA hair Hg reference dose from piscivorous fish consumption, and the fish (tucunaré, pescada) exceeded FDA/EPA 0.5 mg/kg tHg thresholds in at least one season. This strengthens the subsistence-fishing-mehg-disparity synthesis page (already promoted) and the Amazon river system documentation. It is mercury-only. Multi-metal criterion still unmet; ASGM-plus-Pb/Cd paper still needed for promotion.

Three EFSA food additive opinions — pattern noted but not synthesis-worthy

Batch 14 yielded three EFSA opinions (E 422 glycerol, E 475 polyglycerol esters, E 476 PGPR) all recommending tightening EU heavy metal specification limits for food additives. These are low-exposure-contribution items individually and are best filed as sources for the EFSA and EU regulations pages rather than a synthesis thread.

Wade 2022 cacao cadmium — strong ingredient-page update, not a new thread

wade2022-cacao-cadmium-drivers (489 site-years, 10 countries, soil Cd + pH as sole predictors) is a definitive A-tier synthesis paper for the cacao Cd contamination profile. It belongs on ingredients/cacao.md as a primary contributing source and on the mining-impacted thread as context, but does not seed a new synthesis page — the cacao Cd contamination story is documented by existing sources and this paper refines the mechanism.