WoT

Life Sciences Classification: Navigating Pharma, Clinical, and Biotech Taxonomies

April 2, 2026Ram Katamaraja
life-sciencespharmabiotechdeep-dive

Life Sciences Classification: Navigating Pharma, Clinical, and Biotech Taxonomies

TL;DR: Life sciences is the most classification-intensive industry in the world. A single drug touches ATC, NDC, ICD, LOINC, NCI Thesaurus, and FDA regulatory codes before it reaches a patient. WorldOfTaxonomy connects 50+ life sciences systems so clinical, pharmaceutical, and research teams can search and translate across all of them.


The classification landscape

graph TB
  subgraph Clinical["Clinical Coding"]
    ICD10CM["ICD-10-CM\n97,606 codes"]
    ICD10PCS["ICD-10-PCS\n79,987 codes"]
    ICD11["ICD-11 MMS\n37,052 codes"]
    ICDO3["ICD-O-3\n115 codes"]
    DSM5["DSM-5"]
  end
  subgraph Pharma["Pharmaceuticals"]
    ATC["ATC\n6,440 codes"]
    NDC["NDC\n112,077 codes"]
    RXNORM["RxNorm"]
    DEA["DEA Schedules"]
  end
  subgraph Lab["Diagnostics & Lab"]
    LOINC["LOINC\n102,751 codes"]
    SNOMED["SNOMED CT"]
  end
  subgraph Research["Oncology & Research"]
    NCI["NCI Thesaurus\n211,072 codes"]
    MESH["MeSH\n31,124 codes"]
    CTCAE["CTCAE"]
  end
  ICD10CM -.-> ICD11
  ICD10CM -.-> NCI
  ATC -.-> NDC
  LOINC -.-> SNOMED
  NCI -.-> MESH
  ICD10CM -.-> LOINC

One drug, six systems

A single medication can carry codes from six different systems simultaneously. Translating between them is essential for formulary management, drug interaction checking, and regulatory reporting.

graph LR
  DRUG["Paracetamol\n(Tylenol)"] --> ATC2["ATC N02BE01\n(analgesic class)"]
  DRUG --> NDC2["NDC 50580-0449-01\n(specific product)"]
  DRUG --> RX["RxNorm 161\n(normalized name)"]
  DRUG --> WHO2["WHO Essential\nMedicine"]
  DRUG --> ICD2["ICD-10-CM codes\n(diagnoses it treats)"]
  DRUG --> LOINC2["LOINC codes\n(lab tests that monitor it)"]
  style DRUG fill:#E11D48,color:#fff

Clinical coding systems

System Codes Classifies
ICD-10-CM 97,606 Diagnoses (US clinical)
ICD-10-PCS 79,987 Inpatient procedures (US)
ICD-11 MMS 37,052 Diagnoses (WHO global, latest)
ICD-O-3 115 Oncology morphology and topography
ICF 34 Functioning, disability, and health
DSM-5 skeleton Mental health diagnoses
ICPC-2 18 Primary care encounters

Every patient encounter, every claim, every quality metric starts with a diagnosis code.

Pharmaceutical systems

System Codes Classifies
ATC WHO 2021 6,440 Drugs by therapeutic class
NDC 112,077 Drug products by manufacturer and package
RxNorm skeleton Normalized drug names
DEA Schedules 25 Controlled substance classification
WHO Essential Medicines 27 Priority drug list
EDQM Dosage Forms 17 Standard dosage form terms (EU)

Diagnostics and lab

System Codes Classifies
LOINC 102,751 Lab tests and clinical observations
SNOMED CT skeleton Clinical terminology (concepts, relationships)

LOINC covers everything from basic blood panels to specialized genetic assays. Every lab result that flows through an EHR system should have a LOINC code.

Oncology and research

System Codes Classifies
NCI Thesaurus 211,072 Cancer research ontology
MeSH 31,124 Medical literature indexing (PubMed)
CTCAE 27 Adverse event grading in clinical trials
GBD Cause List 23 Global Burden of Disease causes

NCI Thesaurus is the largest single system in WorldOfTaxonomy at 211,072 codes - the reference ontology for cancer research, clinical trials, and FDA/EMA regulatory submissions.

Medical devices

System Codes Classifies
GMDN 17 Global Medical Device Nomenclature
MDR (EU) 22 EU Medical Device Regulation categories
IVDR (EU) 17 EU In Vitro Diagnostics Regulation
FDA 21 CFR 24 US device regulation categories

Procedures and billing

System Codes Classifies
CPT skeleton Outpatient procedures (AMA)
HCPCS Level II 59 Non-physician services and supplies
MS-DRG 50 Medicare inpatient payment groups
G-DRG 26 German inpatient payment groups
NUCC HCPT 94 Healthcare provider taxonomy

The connection problem

The fragmentation is not just inconvenient - it is dangerous. If connections between diagnosis, lab, drug, and procedure codes are broken or outdated, clinical decision support systems fail.

graph LR
  DX["Diagnosis\n(ICD-10-CM)"] --> LAB["Lab Test\n(LOINC)"]
  DX --> RX2["Drug\n(ATC/NDC)"]
  DX --> PROC["Procedure\n(CPT)"]
  RX2 --> AE["Adverse Events\n(CTCAE)"]
  AE --> LIT["Literature\n(MeSH)"]
  DX --> LIT

A clinical decision support system needs diagnosis -> lab test -> drug -> procedure connections. A pharmacovigilance team needs drug -> adverse event -> diagnosis -> literature connections. Each system is maintained by a different organization with a different update cycle.

Health regulation systems

System What It Covers
HIPAA Patient data privacy (US)
FDA 21 CFR Drug and device regulation (US)
MDR / IVDR Medical devices and diagnostics (EU)
CLIA Clinical laboratory standards (US)
Joint Commission Hospital accreditation
CAP Pathology lab accreditation
HITRUST Health information security

Regulatory frameworks are classification systems too. HIPAA provisions map to data handling requirements. FDA 21 CFR parts map to product categories. They are in the graph because compliance teams need to cross-reference regulations with the clinical and pharmaceutical systems they govern.

Using the API for life sciences

Search across all clinical systems

curl "https://wot.aixcelerator.ai/api/v1/search?q=breast+cancer&grouped=true"

Returns codes from ICD-10-CM, ICD-11, NCI Thesaurus, MeSH, ICD-O-3, and domain systems.

Translate between coding systems

curl "https://wot.aixcelerator.ai/api/v1/systems/icd10cm/nodes/C50/translations"

Navigate deep hierarchies

curl "https://wot.aixcelerator.ai/api/v1/systems/icd10cm/nodes/C50.911/ancestors"
graph TD
  ROOT["ICD-10-CM"] --> CH2["Chapter 2: Neoplasms"]
  CH2 --> BLK["C00-C75: Malignant neoplasms"]
  BLK --> C50["C50: Breast cancer"]
  C50 --> C509["C50.9: Breast, unspecified"]
  C509 --> C50911["C50.911: Right breast"]
  style C50911 fill:#E11D48,color:#fff

Find coverage gaps

curl "https://wot.aixcelerator.ai/api/v1/diff?a=icd10cm&b=icd11_mms"

For AI in healthcare

Instead of the AI guessing that "metformin" is "probably A10BA02" in ATC, it calls a tool, gets the verified code, and can navigate the hierarchy to find related drugs, check NDC codes, and cross-reference with ICD diagnosis codes.

The MCP server gives AI agents structured access to all life sciences systems. Clinical decision support, coding assistance, drug interaction checking, research tools - all backed by verified data from authoritative sources rather than LLM hallucination.

The life sciences classification landscape is complex. WorldOfTaxonomy does not simplify it - it connects it.