Therapeutic antibody landscape
Antibodies as medicines: what the molecule is, which format it takes, how it travels from a target idea to a label, and the vocabulary used along the way. This is the business-and-orientation skill; the bench, the literature and the sequence data each have their own.
Volatile facts get looked up
Sort every factual claim into durable or volatile before answering.
- Durable — immunology, molecular formats, mechanisms of action, assay principles, the shape of a regulatory route, what a term means. Answer from knowledge.
- Volatile — how many antibodies are approved, which ones, approval dates, who owns or licensed an asset, what sits in which trial phase, deal values, headcount, who acquired whom. Look these up every time and give the retrieval date alongside the number.
Where to look, by question:
| Question | Source |
|---|---|
| Approved or under regulatory review | The Antibody Society's table (no login), then FDA Drugs@FDA / Purple Book, EMA medicine pages |
| Any antibody therapeutic, approved or not | YAbS, the Antibody Society's searchable therapeutics database (over 2,900 molecules) |
| Near-term approvals and pipeline sweep | the current year's "Antibodies to watch" review in mAbs (for example Crescioli et al., 2026) |
| Trials, phase, enrolment | ClinicalTrials.gov (API v2), EU CTIS; add WHO ICTRP for a global sweep beyond US/EU registries |
| What a name's stem or infix means | WHO INN stem book and the current mAb nomenclature scheme |
| Who owns it, what it costs, what is next | the company's own investor relations deck and annual report |
| Structures and therapeutic sequences | SAbDab, Thera-SAbDab and IMGT/mAb-DB (see antibody-literature-search) |
When no lookup is possible in the session, name the facts left unverified instead of supplying plausible ones. A remembered approval count is wrong often enough to be worthless, and a fabricated one is worse.
Characterise any antibody drug
Five questions. Answer all five and the asset is understood; mark explicitly any that stay unknown.
- Target and epitope — which antigen, which domain or conformation, and whether binding is meant to block, agonise, deplete, bridge or deliver.
- Format — subclass, fragment, bispecific, conjugate, and what that format buys. See references/formats.md.
- Mechanism of action — the causal chain from binding to clinical effect: receptor blockade, agonism, effector recruitment (ADCC, ADCP, CDC), payload delivery, T-cell redirection, protein bridging, receptor internalisation.
- Indication, line and population — the disease, the position in the treatment sequence, and any biomarker gate on eligibility.
- Differentiation — what else addresses that target or indication, and what distinguishes this one: potency, half-life, route and dosing, safety profile, manufacturability, price.
Reference
- references/formats.md — immunoglobulin anatomy, IgG subclasses, fragments, VHH as a drug, bispecific architectures, ADC anatomy, Fc engineering, and decoding a drug's name.
- references/glossary.md — working vocabulary grouped by where it shows up: binding, discovery, developability, PK and effector function, clinical, regulatory, CMC, commercial.
- references/development-path.md — target to label, regulatory routes in the US, EU and Norway, biosimilars, and what kills programmes.
- references/value-and-pitch.md — user fit and path to value for a design tool or result: who buys what, where the value sits, what is defensible, a clinical-problem-to-property table, nanobody selling points.
- references/careers.md — reading a biotech job ad, role families, the IT and data angle, employer research, interview prep.