BindCraft

Peer-reviewed Prospective flag

Summary card

Поле	Значение
Год / version	2025 / NR
Last verified	2026-08-13
Что генерирует	Binder sequence and designed target–binder complex
Форматы	Small non-antibody protein binders
De novo или optimization	De novo binder
DOI	10.1038/s41586-025-09429-6
Paper	https://doi.org/10.1038/s41586-025-09429-6
GitHub / repo	https://github.com/martinpacesa/BindCraft
Service / docs	NR
Лицензия	MIT for repository; PyRosetta/Rosetta terms require separate commercial review
Доступность / commercial-use	Open pipeline; dependency constraints

Задача и generated object

Used in these tasks: De novo non-antibody binder design, Target-conditioned binder generation

Part of these workflows: Target-conditioned miniprotein binder (BindCraft)

Поле	Значение
Required input	Target structure and selected target surface/hotspots
Conditioning	Target chain, hotspot/interface residues, binder length and filter thresholds
Fixed	NR
Editable	NR
Generated	Binder sequence and designed target–binder complex
Single mutations	No
Combinatorial mutations	Full sequence
Multi-objective	Filter-based multi-criterion selection, not Pareto search

Architecture and operating principle

Author-reported (matrix)

Поле	Значение
Base architecture	NR
Model type	5 Target-conditioned binder generator / workflow
Representation	NR
Generation unit	Binder sequence and designed target–binder complex
Conditioning	Target chain, hotspot/interface residues, binder length and filter thresholds
Output	Binder sequence and designed target–binder complex
Sampling	NR
Internal scorer	AF2 confidence/interface metrics + Rosetta scores
Structure generation	AF2 is both differentiable generator component and internal structure evaluator

Input representation

Поле	Значение
Node features	NR
Edge / pair features	NR

Encoder и decoder

Поле	Значение
Encoder	NR
Decoder	NR

Latent or generative state

Conditioning mechanism

Target chain, hotspot/interface residues, binder length and filter thresholds

Generation mechanism

AlphaFold2 backpropagation/hallucination + MPNN redesign + PyRosetta filtering

Sampling: NR

Training objectives

Losses: NR

Mathematical details: NR

Integrated evaluation

AF2 confidence/interface metrics + Rosetta scores

Architectural limitations

Not antibody-specific.
target structure/hotspot selection and AF2 optimization bias matter.
integrated confidence is not experimental affinity.

Uses these architecture components: Inverse folding, Multi-objective optimization

Generator versus evaluation

Component	Role	Architecture class	Integrated or external	Required or optional	Influences generation or post hoc	Runtime included or excluded
AF2-gradient hallucination	primary_generator	AlphaFold2 backpropagation/hallucination + MPNN redesign + PyRosetta filtering	integrated	required	generation	included
AF2 confidence/interface metrics + Rosetta scores	integrated_scorer	NR	integrated	optional	post_hoc	unknown
AF2 is both differentiable generator component and internal structure evaluator	integrated_structure_predictor	NR	integrated	optional	post_hoc	included
Independent complex prediction and binding/functional assays	mandatory_external_dependency	NR	external	required	post_hoc	excluded
Experimental binding/function assays	experimental_validation	NR	external	required	post_hoc	excluded

Training data и biases

Поле	Значение
Training data	NR
Preprocessing	NR
Negative examples	NR
Fine-tuning	NR
Label requirements	NR
Training hardware	NR
Training time	NR

Inference hardware и runtime

Поле	Значение
Candidate count	NR
Batch size	NR
Generation steps	NR
Recycles	NR
Size	NR
Reported runtime	NR
Runtime boundary	NR
CPU	NR
GPU	NR
VRAM	NR
RAM	NR
Disk	NR
Databases	NR

Experimental evidence

Поле	Значение
Evidence class	3–4 prospective binding and functional validation
Retrospective	NR
Prospective binding	Nanomolar binders and functional examples; general binders
Prospective function	Nanomolar binders and functional examples; general binders
Developability	NR
Tested design count	Multiple targets; reported 10–100% task-dependent experimental success rates
Success criterion	Nanomolar binders and functional examples; general binders
Independent validation	NR
Summary	Nanomolar binders and functional examples; general binders

Limitations and failure modes

Not antibody-specific.
target structure/hotspot selection and AF2 optimization bias matter.
integrated confidence is not experimental affinity.

Deployment: Limited internal pilot

Reproducibility, license, commercial use

Поле	Значение
Repo	https://github.com/martinpacesa/BindCraft
License	MIT for repository; PyRosetta/Rosetta terms require separate commercial review
Commercial-use status	Open pipeline; dependency constraints

Alternative tools: AlphaProteo, BoltzGen

Complementary: ProteinMPNN

Related: AlphaProteo, BoltzGen, RFdiffusion

Not directly comparable with

HuAbDiffusion, HuDiff, Humatch, RFantibody, tFold System

Сравнение допустимо только внутри совместимых категорий. Runtime и quality AntiFold vs RFantibody нельзя читать как одну шкалу: разные starting points и generated objects.

Depends on these external modules

Primary paper и полный список источников

BindCraft: one-shot design of functional protein binders

Source matrix: landscape xlsx, last verified 2026-08-13.

Review queue

Bibliographic title must be verified against the PDF.

Детальный source-grounded разбор

1. Что именно делает инструмент

[MATRIX]Вход: Target structure and selected target surface/hotspots

[MATRIX]Generated object: Binder sequence and designed target–binder complex

[MATRIX]Фиксируется: NR. Изменяется: NR.

[MATRIX]De novo или parent-based: De novo binder.

2. Архитектура в одном абзаце

[MATRIX]AlphaFold2 backpropagation/hallucination + MPNN redesign + PyRosetta filtering

4. Математическая модель

NRТочные уравнения, losses и sampling solver будут внесены после чтения paper/code. Не подставляйте родовой UCB/transformer/diffusion, если он не подтверждён для этого инструмента.

Поля NR: conditional probability, update equations, sampling algorithm

5. Физическое устройство состояния

[MATRIX]Seed generated state / output: Binder sequence and designed target–binder complex. Structure generation: AF2 is both differentiable generator component and internal structure evaluator.

6. Физико-химическая интерпретация

[INTERPRETATION]Координаты, frames, PLM embeddings или residue tokens не являются расчётом ΔG, электростатики или растворителя, пока это не сделано внешней energy function. Likelihood / pLDDT / docking score ≠ экспериментальный binding.

7. Training

[MATRIX]Данные: NR. Fine-tuning: NR. Negative examples: NR.

8. Inference

[MATRIX]Sampling: NR. Generation steps: NR. Runtime boundary: NR.

10. Почему, по мнению авторов, метод эффективен

[MATRIX]Author-reported summary (матрица): Nanomolar binders and functional examples; general binders

NRCausal support не классифицирован, пока не прочитаны ablation и experimental sections.

12. Ablation analysis

NRAblation в этом выпуске не разобрана.

13. What the model does not represent

Явный растворитель и ионная сила — обычно отсутствуют в генераторе.
Протонирование, гликозилирование, кофакторы — обычно отсутствуют, если не оговорено.
Конформационная энтропия и долгая динамика — не моделируются одним forward pass.
Экспрессия, агрегация, иммуногенность — не равны likelihood генератора.
Истинная binding free energy — не равна internal score.

14. Evidence and limitations

[MATRIX]Evidence class (матрица): 3–4 prospective binding and functional validation. Prospective binding: Nanomolar binders and functional examples; general binders. Function: Nanomolar binders and functional examples; general binders. Developability: NR.

Not antibody-specific.
target structure/hotspot selection and AF2 optimization bias matter.
integrated confidence is not experimental affinity.

15. Reproducibility

NRРепозиторий: https://github.com/martinpacesa/BindCraft. Лицензия: MIT for repository; PyRosetta/Rosetta terms require separate commercial review. Commit: NR, пока страница не имеет отдельного analysis YAML.

Модульная схема

Модуль	Класс модели	Вход	Выход	Обучается или frozen	Generator или evaluator
AF2-gradient hallucination	AlphaFold2 backpropagation/hallucination + MPNN redesign + PyRosetta filtering	NR	NR	NR	primary_generator
AF2 confidence/interface metrics + Rosetta scores	NR	NR	NR	NR	integrated_scorer
AF2 is both differentiable generator component and internal structure evaluator	NR	NR	NR	NR	integrated_structure_predictor
Independent complex prediction and binding/functional assays	NR	NR	NR	NR	mandatory_external_dependency
Experimental binding/function assays	NR	NR	NR	NR	experimental_validation

Generator versus evaluator

Компонент	Роль	Integrated или external	Required или optional	Влияет на generation или только ранжирует после
AF2-gradient hallucination	primary_generator	integrated	required	generation
AF2 confidence/interface metrics + Rosetta scores	integrated_scorer	integrated	optional	post_hoc
AF2 is both differentiable generator component and internal structure evaluator	integrated_structure_predictor	integrated	optional	post_hoc
Independent complex prediction and binding/functional assays	mandatory_external_dependency	external	required	post_hoc
Experimental binding/function assays	experimental_validation	external	required	post_hoc

Claim–evidence matrix

Утверждение	Предполагаемый механизм	Тип evidence	Датасет / эксперимент	Comparator	Метрика / результат	Источник	Оговорка
Nanomolar binders and functional examples; general binders	AlphaFold2 backpropagation/hallucination + MPNN redesign + PyRosetta filtering	5. prospective functional validation	Multiple targets; reported 10–100% task-dependent experimental success rates	NR	Nanomolar binders and functional examples; general binders	[MATRIX] landscape matrix 2026-08-13	Не читать как полный разбор primary paper.