Unverified50% confidenceFactExact time
Retro190 is a benchmark dataset commonly used for retrosynthesis planning evaluation, containing 190 target molecules, each with a known synthetic route.
1
Sources
50%
Confidence
Long-term
Relevance
10/30/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedThe core task of retrosynthesis analysis is, given a target molecule, to work backward to feasible synthetic routes; PostEra's approach is to model reaction prediction as a sequence-to-sequence or graph-to-graph problem, training models on large-scale reaction datasets (such as USPTO patent reaction data).64% similarUnverifiedA common combination for landing molecular design is a closed loop of "generative models propose candidates + property models score them + retrosynthesis models assess synthesizability".57% similarUnverifiedRetrosynthesis planning is typically modeled as an AND/OR search tree, where OR nodes represent the target molecule and the different reactions that may produce it, and AND nodes represent a specific reaction whose reactant child nodes must all be solvable simultaneously.54% similarUnverifiedDesignability is a commonly used evaluation criterion in generative protein design: the generated backbone is fed into an independent sequence design model (such as ProteinMPNN), and the predicted sequence is then refolded using a structure prediction tool (such as ESMFold or AlphaFold); if the predicted structure is highly consistent with the original backbone (usually judged by an scRMSD threshold), the backbone is considered designable.52% similarUnverifiedDiffusionGemma复用现成的Gemma系列(约26B)模型权重,通过监督微调和采样器蒸馏与强化学习两步改造完成转换52% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1114649API
curl https://kongchang.com/api/v1/knowledge/claims/1114649MCP
get_claim(id=1114649)