分子相互作用基础模型 Bonbon 展现零样本涌现生物学能力
Emergent Biological Capabilities in a Foundation Model for Molecular Interactions
这篇论文把大模型的涌现现象搬到了药物分子层面：只用序列训练的 Bonbon，不做任务监督就能自动学会找结合位点、判断共价与非共价。速度比共折叠快十万倍，湿实验命中率 61.5% 相当扎实。做药物设计或分子机器学习的都该看看。
Bonbon 是一个分子相互作用基础模型，在约 3.2 万亿 token 的蛋白-配体配对序列上训练，无需任务监督、不使用结构坐标，仅靠单一表示空间即涌现出三项零样本能力：作用机制（MoA）识别、结合位点定位（含正构/变构区分）和共价/非共价作用判别。三项能力均在分布外复合物上保持泛化。
论文在 42M 到 1.6B 共六个参数规模上做了对照缩放实验：预训练损失平滑下降，但所有涌现能力都在最大规模上出现跳变，说明基于损失的缩放定律无法预测零样本能力何时出现。训练采用 1:2000 的参数- token 比以适应配对生物数据的稀疏性。
同一冻结表示在结合与亲和力预测上达到 SOTA，速度比共折叠方法快五个数量级。前瞻性湿实验覆盖 7 个蛋白家族的 9 个疾病靶点，从头设计化合物的剂量响应命中率 61.5%，其中包含此前未见报道的全新抑制剂。
Emergent Biological Capabilities in a Foundation Model for Molecular Interactions
Foundation models trained on large unlabeled corpora develop emergent capabilities, tasks that go beyond the training objective. We extend this paradigm to multimodal biology with Bonbon, a foundation model for molecular interactions trained on approximately 3.2 trillion tokens of paired protein-ligand sequences. From a single representation space, Bonbon exhibits three zero-shot emergent capabilities at hierarchical levels of resolution: functional (mechanism of action), residue (binding site localization, including orthosteric/allosteric discrimination), and bond (covalent versus non-covalent engagement). All three capabilities generalize to out-of-distribution complexes. Bonbon was trained on protein-ligand sequences alone, without any task-specific supervision and without structural coordinates. We conducted controlled scaling experiments across six model scales from 42M to 1.6B parameters; every capability evaluated shows a sharp jump at the largest scale on continuous metrics, even as pretraining loss decreases smoothly. This demonstrates that loss-based scaling laws alone are insufficient to predict when zero-shot emergent capabilities become available. The capabilities reported here emerge after training at a 1:2,000 parameter-to-token ratio to accommodate the sparsity of paired biological interaction data. The same frozen representations achieve state-of-the-art binding and affinity prediction at five orders of magnitude greater speed than co-folding methods. In prospective wet-lab validation across nine disease targets spanning seven protein families, Bonbon achieves a 61.5% dose-response hit rate on de novo compounds, yielding structurally novel molecules not previously reported to inhibit these proteins, including for out-of-distribution targets.
本条情报归属 AITOP「论文」核心栏目，重点聚焦北美前沿实验室与硅谷科技巨头。在跨区域产业观察坐标中，该动态直接关联「AI 开发者工具链、Coding Agent 与工程化实践」研究主线，反映出当前技术演进与商业落地的最新风向。涉及核心实体与技术议题包括：Bonbon、分子相互作用、缩放定律、药物发现，读者可通过下方标签进入对应专题页查阅历史脉络与关联对比。
信源核验说明：本文原始线索采编自「arXiv cs.LG」（@Adi Mashiach），经 AITOP 多维评分管线结构化提炼。