AI for Bio · Model

Claude protein-design study (general model + public tools)

A general language model orchestrates public tools to design binders, tested by outside labs

ProteinsMixedUnited States, SwitzerlandProduct only, no public artifact中文

Overview

NameClaude protein-design study (general model + public tools)
Full nameStudy of a general language model acting as a design agent over public protein-design and co-folding tools
Ambiguous names (not auto-linked)Claude
FamilyNot a family entry. The subject is a study configuration, not a model family; the same partners' separate 2025-10 study is listed on its own
OrganizationAnthropic, Adaptyv Bio, Twist Bioscience
Organization typeMixed
CountryUnited States, Switzerland
Model classAgent (general model orchestrating public design tools)
DomainProteins (also: Text, LLMs & agents, Molecule generation)
First release
Basis: Developer research post published (a partner case study followed the next day) source
Other milestones
  • First public evidence: · Developer research post source
  • Official announcement: · Developer research post, releasing prompts, computational models and in vitro data source
Latest update · no later event
Basis: Partner laboratory case-study report published
The latest sourced event is the partner laboratory's case-study report; a joint competition ran 2026-09-28 to 10-31 with results due 2026-12-15 and nothing citable yet source
The most recent sourced event, not a claim that nothing changed since.
StatusProduct only, no public artifact · No public model artefact: the general model is closed and what is public is prompts and datasets
OpennessCode: Closed
Weights: Closed
The general model is closed and no code was released; prompts, computational models of the designed complexes and all in vitro and computational data are public as a dataset
Benchmarks reported1 counted (retrospective 1)
Wet-lab validation2 recorded campaigns
Unverified fields† training_data.a_public, training_data.b_private_licensed, background.funding, other_evaluations[0]
As of2026-11-16

What it does

Not a new generative network: a general language model acts as an agent, orchestrating public backbone-generation, sequence-design and co-folding filtering tools per target, after which two outside organisations express the designs and test binding. Unlike BindCraft or BoltzGen the contribution is orchestration and selection; the tools called include RFdiffusion2/3, BoltzGen, Proteina-Complexa, Protenix and ESMFold2, though the tool list appears only in second-hand summaries.

Benchmarks reported

Every row is one pairing from the survey's relation file; a score belongs to the benchmark page's ranking, where it is shown with metric, eval set, source type and date. 1 of 1 pairings are counted.

BenchmarkRelationEval set / protocolSource typeNote and evidence
28 Adaptyv competitions / BenchBBRetrospective dataset use—Third-partyOrganizer-run anonymous external test on 16 prior-competition targets; not a public competition · evidence

Other evaluations (outside the 44)

  • Per-target and per-configuration hit rates (Own task set) Third-party: Both parties report hit rates per target and per design configuration with wide variation; some per-target figures appear only in second-hand coverage
  • The same partners' separate 2025-10 binder study (Case study) Self-reported: An earlier study applied the same approach to a different target class and sent designs for testing; it is a separate study and its figures are not merged here

Developers report: a general model can autonomously orchestrate public tools through the whole design pipeline, yielding binding designs for most targets at a high expression rate.

Independent evidence: No independent evidence found. The wet lab was executed and compiled by an interested partner that also chose the targets; no disinterested third-party evaluation was found

Training data

Group in Figure 7: Public data + wet lab for validation only.

ClassValueDetail
a. Public databasesNot disclosedThe general model's training data is not public; each public tool it calls has its own training data source
b. Private or licensed dataNot disclosed
c. Own wet-lab data used for trainingNoNo dedicated training in this study; wet-lab results were not fed back source
d. Wet lab used only for validationYesTwo outside organisations handled expression and binding assays, for validation only source

Wet-lab validation

Every rate carries its numerator and denominator. Claims that could not be verified are shown as unverified, without numbers; different paper versions are separate entries.

Partner-run external test (blinded submission) (dataset + partner report)
Targets 15 · Designs tested 1320 · Hits 354 · Success 354/1,320 (final call pooled across two vendors; one vendor's raw calls give 336); 22.6%-35.1% by configuration · Best affinity —

Denominator basis: 1,440 designs on 16 targets, of which 15 targets and 1,320 designs have usable data (one target's 120 designs excluded for aggregation and uncertain data); expression 1,235 of 1,296 tested (95.3%)
Attributing 354 to a single vendor is inaccurate: it is the pooled two-vendor final call, and per-configuration rates must be listed separately rather than merged
Executor: Executed by a partner laboratory and a gene-synthesis vendor; the organiser is an interested party, so this is not an independent evaluation

Third-party · · verified · source
Affinity comparison with past competition winners (partner report)
Targets 6 · Designs tested — · Hits — · Success best designs showed greater affinity than five of the six past winners · Best affinity —

Denominator basis: Best design compared with the winning binder of each of six past competitions
The published "would have won five of six competitions" overstates the source: it is a best-binder affinity comparison, not a ranking simulation, and one competition's best binder was not beaten
Executor: Executed and compiled by the partner laboratory

Third-party · · verified · source

Background

  • Organization: The design side is a US public-benefit AI company founded in 2021 and based in San Francisco (the work is credited to its science team, with no individuals listed); the wet-lab side is a European protein measurement service company and a US gene-synthesis company, with another vendor supplying a high-throughput surface plasmon resonance platform.
  • Funding: The design side announced a $30B Series G in 2026-02 at a $380B post-money valuation, co-led by several investors (second-hand reporting).
  • Products and partnerships: The products are the company's general model family and its science programme; with the wet-lab partner it also ran a joint competition 2026-09-28 to 10-31, with results due 2026-12-15.
  • Related models: BindCraft, BoltzGen, RFdiffusion2 / RFdiffusion3, Proteina-Complexa

Reading notes

  1. The comparison is against historical competition data rather than a matched control, and the interested organiser chose the targets.
  2. The pooled hit rate swings widely with target difficulty, including a zero-hit target, so it is not a stable rate.
  3. Only binding was measured; there are no functional, specificity or developability conclusions.
  4. The tool list exists only in second-hand summaries, so the gain cannot be split between the orchestration and the tools called.

Caveats

  1. The final hit call pools two vendors; one vendor's raw calls differ.
  2. The "would have won five of six" framing is wrong; correctly, the best designs out-bound five of the winners by affinity.
  3. The entry is retrospective dataset use and does not count toward reported benchmarks.
  4. The joint competition's results are not out and cannot be cited.

Sources

  1. Developer research post (company blog)
  2. Partner laboratory case study (company blog)
  3. Public dataset (prompts, computational models, in vitro data) (official site)
  4. Second-hand summary of the tool list and per-target figures (second-hand) second-hand
  5. Assay platform vendor coverage (press)

Corrections

2 corrections
FieldWasNowBasisFound
wetlab_validation354 binders measured by a single vendor; one hit rate of 26.8%354/1,320 is the final call combining two vendors (336 for a single vendor); expression 1,235/1,296 = 95.3%; per-configuration 22.6%–35.1% listed separatelysource2026-11-16
wetlab_validationCould have won 5 of 6 competitionsThe best design's affinity beats 5 of the 6 winners, and does not exceed the best binder of one of themsource2026-11-16

Part of the AI for Bio benchmarks survey · all 112 models · open in the explorer. Record as of 2026-11-16.