EUTECNIQ Sciences

See every page on one mapArrow keys move through results · Enter opens · Esc closes

Platform · The data asset Primary sources Proposed dataset

The data asset

Open databases catalog structures and the organisms they come from, but none of the five below is organized around an authenticated physical lot. EUTECNIQ proposes to build that record itself: one lot, measured end to end, kept as signed and tamper-evident data.

695,133unique natural-product structures in COCONUT 2.0, from 63 sources: structures, not lot measurements
119 millioncompounds in PubChem (2025 update), with 295 million bioactivities
750,000+referenced structure–organism pairs in LOTUS, released under CC0
Jul 20, 2028FSMA 204 traceability compliance date, for foods on the Food Traceability List

Status. EUTECNIQ is a company in formation (what exists today). It holds no dataset, model, ledger, laboratory or client data. Everything about EUTECNIQ on this page is a proposal.

Chandrasekhar et al., Nucleic Acids Res 2025;53(D1):D634–D643; Kim et al., Nucleic Acids Res 2025;53(D1):D1516–D1525; Rutz et al., eLife 2022;11:e70780; Public Law 119-37.

Related, not repeated here: Digital lab and traceability (Part 11, ALCOA+, lot genealogy, FSMA 204, a live hash-chain demonstration) · Large Botanical Model (architecture, open-data matrix, scaling) · Botanical Intelligence (the first narrow model) · Computational science and AI (provenance fields, data rights) · Model register

Three-dimensional rendering of a stack of six colored glass layers, labeled identity, chemistry, contaminants, process history, stability and biology, under one lot tag.
  1. 1 One authenticated lot
  2. 2 Six layers, measured the same way for every lot

The proposed lot record, rendered by EUTECNIQ. No lot data exist yet.

Public data

What public data hold

Open databases are large and useful for chemistry. Each is organized around a compound, a structure–organism pair or a food, and each states its own terms. COCONUT states three different licenses in three places.

None of the five below ties its values to a voucher-authenticated botanical lot carrying that lot's own fingerprint, contaminant results and bioassays. That pairing is what an identity or quality model needs, and it has to be measured.

Five open databases: steward, what one record is organized around, size as stated by the provider, and terms as stated
DatabaseOne record is organized aroundSize, as statedTerms, as stated
PubChem (NCBI, NIH)A compound or a deposited substance, with deposited bioactivity results119 million compounds, 322 million substances, 295 million bioactivities, more than 1,000 data sources (2025 update)NCBI "places no restrictions on the use or distribution"; submitters may claim patent, copyright or other rights in submitted data
ChEMBL 36 (EMBL-EBI)A compound, a target and a bioactivity value2.8 million distinct compounds; 17,803 drug targets (release of October 15, 2025)Creative Commons Attribution-ShareAlike 3.0 Unported: credit required; adaptations distributed under the same license
LOTUS (hosted on Wikidata)A structure, an organism and the reference reporting the pair750,000+ referenced structure–organism pairsCC0
COCONUT 2.0A natural-product structure, with organisms and references695,133 unique structures from 63 sources (September 2024 release)Stated inconsistently: CC0 on the download page; each source's own license on GitHub; CC BY 4.0 on Zenodo
FoodData Central (USDA)A food: Foundation Foods, SR Legacy, FNDDS survey foods, Experimental Foods and Branded FoodsFive data types; Branded Foods updated monthly, FNDDS every two years, SR Legacy final in 2018Public domain, CC0 1.0; USDA asks to be credited as the source

Kim et al., PubChem 2025 update (abstract); NCBI Website and Data Usage Policies; EMBL-EBI, "ChEMBL 36 is live" (Oct 15, 2025) and ChEMBL home page; Rutz et al., eLife 2022; Chandrasekhar et al., NAR 2025 (COCONUT license statements per the v21 research pass); USDA FoodData Central home page; Creative Commons BY-SA 3.0 and CC0 1.0 deeds. All read October 28, 2026. Counts for these and other sources on a log scale: Computational science and AI. Summary, not legal advice.

Proposed dataset

One record per lot

EUTECNIQ proposes to measure each lot it handles the same way and to keep the results in one lot record. Identity, chemistry, contaminants, process history, stability and biology could then be compared across lots and botanicals. Part 111 already requires lot identifiers and genealogy links (21 CFR 111.155(d)(1), 111.260); the proposal would extend the same key to research data.

Six kinds of measurement on the same lot would join one record; external data would enter only with a provenance and license tagProposed design
Proposed lot record: six kinds of measurement on the same physical lot, from voucher and DNA identity through receipt testing, process, characterization, stability and biology, join one lot record; external literature and database records enter only with a provenance and license tag; the lot record feeds release decisions, model training sets and customer verificationMeasured on the same physical lotSource and identityvoucher specimen;DNA barcode(partner); collectionsite and dateReceipt testingHPTLC fingerprint;macroscopy andmicroscopy; elements,pesticides, microbialcountsProcessextraction and dryingparameters; yields;batch recordCharacterizationLC-HRMS fingerprint;quantitative markersStabilityassay and physicaltests at each pull pointBiologyin vitro and in vivoresults; human PKlaterOne lot recordlot ID and genealogy links (21 CFR 111.155(d)(1), 111.260); every value carries unit, method version, instrument ID, analyst and time; changes kept in an audittrail (§ 11.10(e))External recordsliterature, regulators, compendia, opendatabases; enter with a provenance and licensetag (section 3)Release decisions andcertificatesQC approval with e-signatureTraining and test setssplit by botanical and by lotCustomer verificationhash of each certificateSchematic. Proposed record design; no EUTECNIQ data exist. Human PK would be added only if clinical studies are run.Proposed lot record: six kinds of measurement on the same physical lot, from voucher and DNA identity through receipt testing, process, characterization, stability and biology, join one lot record; external literature and database records enter only with a provenance and license tag; the lot record feeds release decisions, model training sets and customer verificationMeasured on the same physical lot1Source and identityvoucher specimen; DNA barcode (partner);collection site and date2Receipt testingHPTLC fingerprint; macroscopy andmicroscopy; elements, pesticides, microbialcounts3Processextraction and drying parameters; yields; batchrecord4CharacterizationLC-HRMS fingerprint; quantitative markers5Stabilityassay and physical tests at each pull point6Biologyin vitro and in vivo results; human PK laterOne lot recordlot ID and genealogy (21 CFR 111.155(d)(1), 111.260);unit, method version, instrument, analyst and timeon every value; audit trail (§ 11.10(e))External recordsenter only with a provenance and license tagRelease decisions and certificatesQC approval with e-signatureTraining and test setssplit by botanical and by lotCustomer verificationhash of each certificateSchematic. Proposed design; no EUTECNIQ data exist.
Schematic, restating the schema table below. Lot-link rules: 21 CFR 111.155(d)(1) and 111.260; audit trail: 21 CFR 11.10(e) (eCFR, current to Oct 21, 2026). How the laboratory would capture each record: Digital lab and traceability.
Proposed lot-record schema: field, unit or format, method standard, source instrument and when it would be available
FieldUnit or formatMethod standardSource instrumentAvailable
Lot ID and genealogyIdentifier; parent and child lot IDs21 CFR 111.155(d)(1); batch record, 111.260LIMSPhase I
Voucher specimenHerbarium accession number; imagesUSP <563> macroscopic and microscopic identificationStereo and compound microscopes; herbarium (partner)Phase I
DNA barcodeSequence; marker region; reference library and versionUSP <563> describes no DNA method; partner's validated methodSequencer (partner laboratory)Partner
HPTLC fingerprintPlate images; RF of zonesUSP <203>HPTLC systemPhase I
LC-HRMS fingerprintm/z, retention time (min), intensity; raw data fileSpecificity per ICH Q2(R2), used as good practiceHigh-resolution LC-MSPhase I
Quantitative markers% w/w or mg/gICH Q2(R2) validation; compendial monograph where one existsHPLC-UV; LC-MS/MSPhase I
Elemental contaminantsµg/g; µg/day at the labeled doseICH Q3D(R2) permitted daily exposuresICP-MSPhase I
Pesticide residuesmg/kgUSP <561> Articles of Botanical OriginGC-MS/MS; LC-MS/MSPhase I
Microbial countscfu/g; specified organisms absent or presentUSP <2023>Incubators; plate countsPhase I
Process parameters°C, bar, min, solvent ratio, yield (%)Batch production record, 21 CFR 111.260Process sensors; balancesPhase I (laboratory) Phase II (pilot)
Stability% of initial assay at each pull point (months); storage conditionICH Q1A(R2) storage conditions, used as good practiceStability chambersPhase I
In vitro resultsIC50 (µM) or % of control, with positive and negative controlsWritten protocol per assayPlate readers; LC-MSPartner
In vivo resultsDose (mg/kg bw/day); study endpointsOECD test guidelines; GLP (21 CFR Part 58) as the study plan requiresContract research laboratoryPartner
Human pharmacokineticsCmax (ng/mL), tmax (h), AUC (ng·h/mL)Approved protocol, informed consent and IRB review (21 CFR Parts 50 and 56 where FDA rules apply)LC-MS/MS bioanalysisOnly if studies are run
External record provenanceSource, identifier, version, retrieval date, licenseAdmission rule (next section)Curator entry, second-person checkPhase I

eCFR Title 21 Parts 11, 50, 56, 58 and 111 (current to Oct 21, 2026); USP <203>, <561>, <563> and <2023> (chapter titles and scope as published); ICH Q1A(R2) (Step 4, Feb 6, 2003), Q2(R2) (adopted Nov 1, 2023), Q3D(R2) (2022). ICH quality guidelines are written for drugs and are cited here as good practice. Phases follow Facilities and roadmap; DNA barcoding is assigned to a partner there.

Inclusion policy

Reputable sources only

External records would be admitted only from primary literature, regulators, compendia and maintained databases, and only after four checks. Each admitted record would carry its source, version, retrieval date and license, and would stay tagged as external so that it is never mistaken for a measurement on an EUTECNIQ lot. Nothing would be scraped from the web.

Four checks before an external record is admitted; failing any one excludes or holds it outProposed policy
Proposed admission rule for external records: four checks in order (source class, license, retraction status, provenance); a record that fails any check is excluded or held out; a record that passes all four is admitted, tagged as external and kept apart from measured lot data1. Source classPeer-reviewed paper,regulator, compendium,or a database with anamed maintainer and apaper describing it?Excluded: scrapedpages, marketing copy,forums, unattributedtablesno2. LicenseDo the stated terms allowthe intended use: internalanalysis, model training,redistribution?Held out, or used onlywhere the terms allowno3. StatusNot retracted;corrections and errataapplied?Excluded; the retractionis logged against thesourceno4. ProvenanceIdentifier, version,retrieval date, license andcitation recorded?Returned for curation; notadmittednoAdmittedtagged 'external'; keptapart from measured lotdatayesProposed policy. A license to download is not automatically a license to train or to redistribute.Proposed admission rule for external records: four checks in order (source class, license, retraction status, provenance); a record that fails any check is excluded or held out; a record that passes all four is admitted, tagged as external and kept apart from measured lot data1. Source classPeer-reviewed paper, regulator, compendium, or adatabase with a named maintainer and a paperdescribing it?No → Excluded: scraped pages, marketing copy,forums, unattributed tables2. LicenseDo the stated terms allow the intended use: internalanalysis, model training, redistribution?No → Held out, or used only where the terms allow3. StatusNot retracted; corrections and errata applied?No → Excluded; the retraction is logged againstthe source4. ProvenanceIdentifier, version, retrieval date, license and citationrecorded?No → Returned for curation; not admittedAll four pass → admittedtagged 'external'; kept apart from measured lot dataProposed policy. A license to download is not a licenseto train.
License terms as stated on the Creative Commons CC0 1.0 and BY-SA 3.0 deeds (read Oct 28, 2026). Retraction status example: Retraction Note, BMC Med 2024;22:279.
Source classes, examples, how each would be admitted and the license note
Source classExamplesAdmitted asTerms to check
Peer-reviewed literatureJournal articles with a DOIData values with the citation and table or figure number, entered by one curator and checked by a secondValues are cited; article text is not redistributed
RegulatorseCFR, FDA, EFSA, EMARule text and limits with section number and retrieval dateCited with the version in force on the retrieval date
CompendiaUSP–NF, European PharmacopoeiaMethod and limit references by chapter or monographSubscription texts: cited, not copied
Maintained databasesPubChem, ChEMBL, LOTUS, COCONUT, FoodData CentralRecords with database name, release and identifierCC0: no permission needed. CC BY-SA: credit, and adaptations shared under the same license. Unclear terms: held out
Client dataResults on contracted samplesOnly with written permission, segregated by contractThe contract governs every use
ExcludedScraped web pages, marketing copy, unattributed tables, unlicensed content, retracted papersNot admitted—

Creative Commons, CC0 1.0 Universal and Attribution-ShareAlike 3.0 Unported deeds; NCBI Website and Data Usage Policies. Proposed policy; summary, not legal advice.

Machine learning

Models the record would serve

The lot record would serve narrow models, each with a metric, a baseline and a threshold fixed before data are seen, tested on botanicals and lots held out of training. The architecture is on the Large Botanical Model page, the first objective on Botanical Intelligence, and the checks on every calculator in the model register.

Proposed machine-learning tasks: input from the lot record, metric, baseline to beat, and what each would not claim
TaskInput from the lot recordMetricBaseline it must beatIt would not claim
Identity and adulteration flagsHPTLC and LC-HRMS fingerprints, with voucher and DNA identity as labelsTop-k accuracy; sensitivity at fixed specificity; precision–recall AUCSpectral library search; PLS discriminant analysis; marker-ratio rulesThat a lot is authentic without the identity test 21 CFR 111.75(a)(1)(i) requires
Marker contentRaw-material fingerprint; process parametersMean absolute error; prediction-interval coverageGradient boosting or Gaussian processes on descriptorsA release value; the validated assay decides
Stability and formulation designComposition, process history, storage condition, pull-point resultsMean absolute error on held-out lots; interval coverageRegression on each lot's own stability dataA shelf life without real-time data
Interaction screeningConstituent profile; in vitro enzyme and transporter resultsPrecision–recall AUC on held-out constituentsLiterature search by a trained scientistA clinical interaction; only human studies show that
Every task—Expected calibration error; abstention rate—Performance outside the botanicals and lots in training

Metrics and baselines follow the Large Botanical Model page (section "Metrics, baselines and thresholds"); identity-test rule: 21 CFR 111.75(a)(1)(i), eCFR. Proposed tasks; no model exists.

Certificates and supply chain

Tamper-evident, not self-proving

Each certificate of analysis and supply-chain event, from harvest to shipment, would be a signed record hashed with the one before, so a later edit is detectable. Part 11 already requires audit trails that do not obscure earlier entries, signatures showing name, time and meaning, and signatures linked to their records (§§ 11.10(e), 11.50, 11.70). FSMA 204 covers only foods on the Food Traceability List, from July 20, 2028.

Each record's hash covers the one before; a customer can check a certificate against the anchored hash, but not what happened before entryProposed design
Proposed tamper-evident record chain from harvest to shipment: each event record is hashed together with the previous record's hash; the latest hash is anchored by a time-stamp authority, a permissioned ledger or a public blockchain; a customer recomputes the hash of the certificate received and compares it with the anchored value. What happened before a record was entered (sampling, substitution, an incorrect result) lies outside the chainEvent records, in the order they happenHarvest or collectiongrower or collector,place, date, voucherreferenceh1Receiptsupplier lot ID,quantity, quarantinestatush2Testing andcertificateresults; signed: name,date and time,meaning (§ 11.50)h3Extraction anddryingbatch record: inputs,parameters, yieldh4Finished-lot releaseQC approval withe-signatureh5Shipmentcustomer, quantity,dateh6Each hash covers its record and the previous hash: hₙ = SHA-256(recordₙ ‖ hₙ₋₁). Editing any earlier record changes every later hash.Anchor the latest hash (design choice, table below)RFC 3161 time-stamp token, a permissioned ledger entry, or a public blockchain transactioncarrying the hash onlyCustomer verificationrecompute the hash of the certificate received;compare with the anchored valueOutside the chain: what was sampled, whether the sample represents the lot, and whether an entered result is correctA ledger cannot tell whether input data reflect real-world events (NIST IR 8202, § 7.3). These are controlled by sampling plans, orthogonal identity tests atreceipt and audits.Schematic. Records and the anchoring method are proposals; EUTECNIQ operates no such system.Proposed tamper-evident record chain from harvest to shipment: each event record is hashed together with the previous record's hash; the latest hash is anchored by a time-stamp authority, a permissioned ledger or a public blockchain; a customer recomputes the hash of the certificate received and compares it with the anchored value. What happened before a record was entered (sampling, substitution, an incorrect result) lies outside the chainEvent records, each hashed with the one beforeHarvest or collectiongrower or collector, place, date, voucherreferenceh1Receiptsupplier lot ID, quantity, quarantine statush2Testing and certificateresults; signed: name, date and time,meaning (§ 11.50)h3Extraction and dryingbatch record: inputs, parameters, yieldh4Finished-lot releaseQC approval with e-signatureh5Shipmentcustomer, quantity, dateh6hₙ = SHA-256(recordₙ ‖ hₙ₋₁): editing an earlier recordchanges every later hash.Anchor the latest hashRFC 3161 time-stamp, permissioned ledger or publicblockchain (hash only)Customer verificationrecompute the certificate's hash; compare with theanchored valueOutside the chainwhat was sampled, whether it represents the lot,whether a result is correct; controlled by samplingplans, orthogonal tests and audits (NIST IR 8202,§ 7.3)Schematic; EUTECNIQ operates no such system.
21 CFR 11.10(e), 11.50, 11.70 (eCFR, current to Oct 21, 2026); NIST IR 8202 (2018), sections 2.2 and 7.3; RFC 3161 (2001). Try a working hash chain: Digital lab and traceability.

A ledger preserves what was entered; it does not make it true. NIST notes there "may be no method to determine if the input data reflects real world events". A 2013 BMC Medicine DNA-barcoding study of herbal products was retracted on July 4, 2024 after a university investigation found evidence of data fabrication. Identity would therefore still rest on sampling and orthogonal tests at receipt.

Four record designs compared on who can write, who can verify, privacy, cost and regulatory fit
DesignWho can writeWho can verifyPrivacy of certificate contentCost (qualitative)Regulatory fit
Hash chain in the LIMS, with RFC 3161 time-stampsEUTECNIQ's systems onlyAnyone given the record and the hash or time-stamp tokenContent stays private; only hashes leaveLowest: software and a time-stamp serviceSmallest system to validate; Part 11 duties unchanged
Permissioned ledgerNamed, authorized participantsParticipants, and auditors if admittedSome platforms reveal content only to the parties involvedModerate: participants run nodes and agree governanceEach participant keeps its own Part 11 duties; shared validation and governance
Public blockchain, hash onlyAnyone; no authorization neededAnyoneContent private; the hash and its timing are publicNetwork transaction feesNot required by any rule; adds dependence on a network EUTECNIQ does not control
Public blockchain, content on chainAnyoneAnyoneResults, suppliers and clients exposed to anyoneNetwork transaction feesNot proposed: conflicts with client and supplier confidentiality

NIST IR 8202, Blockchain Technology Overview (Oct 2018): in a permissionless network "anyone can read and write to the blockchain without authorization"; permissioned networks "limit participation", and some can "selectively reveal transaction information" (executive summary; section 2.2). RFC 3161, Time-Stamp Protocol (Aug 2001). Retraction Note, BMC Med 2024;22:279, doi:10.1186/s12916-024-03504-x. Public Law 119-37. Conceptual comparison: no costs or performance figures are claimed, and the choice is open.

Use of the data

Where the value would come from

The same records would serve four uses. None has a customer, price or contract. The dataset would grow with the facility phases, in units such as lots, spectra and pull points; no counts are projected, because they depend on contracts and gates.

Four proposed uses of the data: who would use it, what would be delivered, and the condition
UseWho would use itWhat would be deliveredCondition
Internal decisionsEUTECNIQ quality and R&DEach new lot compared with every earlier lot of the same botanical, for supplier qualification, release and formulationOperates inside the quality system
Services to clientsIngredient suppliers and brandsIdentity and quality data packages; model-assisted screening reported with its metric and limitsValidated methods; no ISO/IEC 17025 accreditation is held
LicensingIngredient and software firmsDe-identified datasets or trained models built from EUTECNIQ's own measurementsOwn data only; client data only with written permission; external data only where its terms allow
Certificate verificationBuyers of clients' ingredientsA check that a certificate matches its anchored hash and is unchanged since signingVerifies integrity, not the truth of the results
What each phase would add to the record, in units rather than countsProposed phasing
Proposed growth of the dataset by phase, in units rather than counts: Phase I adds received-lot identity records, contaminant results, LC-HRMS spectra and marker results, laboratory process runs, stability results and contracted client results; Phase II adds pilot batch records and sensor time series; Phase III adds commercial batch and distribution records; DNA barcodes, in vitro and in vivo results and human pharmacokinetics would come from partner studies not tied to a phasePhase Ilaboratory corePhase IIpilot, after Gate 1Phase IIIcommercial, after Gate 2Partner studiesnot tied to a phaseReceived lots: voucher reference, HPTLC imageReceived lots: voucher reference, HPTLC image: from Phase I, laboratory coreContaminants: elements, pesticides, microbesContaminants: elements, pesticides, microbes: from Phase I, laboratory coreLC-HRMS spectra and marker resultsLC-HRMS spectra and marker results: from Phase I, laboratory coreLaboratory extraction, formulation, drying runsLaboratory extraction, formulation, drying runs: from Phase I, laboratory coreStability results per pull pointStability results per pull point: from Phase I, laboratory coreClient results, under written agreement onlyClient results, under written agreement only: from Phase I, laboratory corePilot batch records and sensor time seriesPilot batch records and sensor time series: from Phase II, pilot, after Gate 1Commercial batch and distribution recordsCommercial batch and distribution records: from Phase III, commercial, after Gate 2DNA barcodes (partner laboratory)DNA barcodes (partner laboratory): from Partner studies, not tied to a phaseIn vitro and in vivo study resultsIn vitro and in vivo study results: from Partner studies, not tied to a phaseHuman PK, if clinical studies are runHuman PK, if clinical studies are run: from Partner studies, not tied to a phaseBars run on into later phases: the record accumulates. No counts are projected; they depend on contracts and gates.Proposed growth of the dataset by phase, in units rather than counts: Phase I adds received-lot identity records, contaminant results, LC-HRMS spectra and marker results, laboratory process runs, stability results and contracted client results; Phase II adds pilot batch records and sensor time series; Phase III adds commercial batch and distribution records; DNA barcodes, in vitro and in vivo results and human pharmacokinetics would come from partner studies not tied to a phasePhase I · laboratory coreReceived lots: voucher reference, HPTLC imageContaminants: elements, pesticides, microbesLC-HRMS spectra and marker resultsLaboratory extraction, formulation, drying runsStability results per pull pointClient results, under written agreement onlyPhase II · pilot, after Gate 1Pilot batch records and sensor time seriesPhase III · commercial, after Gate 2Commercial batch and distribution recordsPartner studies · not tied to a phaseDNA barcodes (partner laboratory)In vitro and in vivo study resultsHuman PK, if clinical studies are runEach phase keeps what earlier phases added. Nocounts are projected.
Phases and gates as on Facilities and roadmap; partner work as assigned there. What EUTECNIQ would and would not deliver: Deliverables; how the model compares with suppliers and laboratories: How EUTECNIQ differs.
Risks

What could go wrong

A dataset is worth only as much as its weakest label and its narrowest license. Seven risks, how each would show, and the proposed control.

Seven risks to the proposed data asset, how each would show, and the proposed control
RiskHow it would showProposed control
Data qualityMislabeled reference lots; methods changing between years; models learn the errorVoucher and orthogonal identity on every reference lot; method version on every value; re-analysis rules
Too few lotsModels cannot beat their baselinesStop rules set in advance; no model output used until it beats its baseline (Botanical Intelligence)
LicenseExternal terms forbid training, or share-alike terms reach a derived datasetLicense field on every record; unclear terms held out; external records kept apart
Client confidentiality and privacyClient results or supplier names exposed through a shared model or a public ledgerSegregation by contract; only hashes leave the system; no content on public chains; human data only under consent and IRB review
Model overclaimingAccuracy reported on random splits; claims outside the training domainSplits by botanical and lot; thresholds set in advance; abstention; model cards
Regulatory acceptanceFDA or customers do not accept a model output in place of a testModels propose and validated tests decide; the Part 111 identity test stays (21 CFR 111.75(a)(1)(i))
Ledger misread as proofAn anchored certificate taken as proof that the material is authenticEvery verification result states that it confirms integrity, not truth

Proposed controls. Related: Computational science and AI (data rights, human review); Quality system.

Where EUTECNIQ would work

Where EUTECNIQ would work Proposed

EUTECNIQ would measure each lot it handles once, completely, and keep the result as a signed, linked record. The models and the verification service would rest on that record. Neither would replace the identity test, the validated assay or the human study.

Working on botanical data, certificates or traceability?

Discuss a data, modeling or verification collaboration.

Health information notice. These statements have not been evaluated by the Food and Drug Administration. Nothing on this site is intended to diagnose, treat, cure or prevent any disease, and nothing here is medical advice.

Read the full notice

The information is for general education only. It does not replace advice, diagnosis or treatment from a qualified health professional, or any medication a clinician has prescribed. Do not start, stop or change any medicine, supplement or herbal product because of something you read here. Talk to your doctor or pharmacist first, especially if you are pregnant or breastfeeding, have a medical condition, take any medicines, are preparing for surgery, or are caring for a child. Research findings, traditional uses and safety data are reported as their sources state them; they are not recommendations or claims about any product. If you think you may have a medical emergency, call your doctor or 911 at once.

EUTECNIQ Sciences sells no supplements or health products today and receives nothing from any product or ingredient mentioned on this site; any commercial interest it later has in an ingredient will be stated on that ingredient’s page.

Ready to start a development project?

Tell us about an ingredient you want to specify, verify or license. Questions? Email [email protected]