EnzyProp model checkpoints for enzyme property prediction

This record provides the pretrained model checkpoints for EnzyProp, a structure-aware toolkit for enzyme property prediction.

EnzyProp predicts three enzyme properties from protein sequence and PDB structure information:- topt: optimal temperature- phopt: optimal pH- pi: isoelectric point

Files included in this record:- topt_best_model.pt- phopt_best_model.pt- pi_best_model.pt

Source code and documentation are available on GitHub:https://github.com/LiuSH0402/EnzyProp

Note:- These files are the trained EnzyProp checkpoints.- HuggingFace base models such as ESM2 and ProtBert are not included in this record.- The hf_cache directory is generated locally during model download and use.

本数据集收录了EnzyProp的预训练模型检查点，EnzyProp是一款面向酶性质预测的结构感知工具包。EnzyProp可基于蛋白质序列与PDB（Protein Data Bank）结构信息，预测三类酶学性质：topt为最优温度，phopt为最适pH，pi为等电点。
本数据集包含以下文件：
- topt_best_model.pt
- phopt_best_model.pt
- pi_best_model.pt
源代码与相关文档可于GitHub平台获取，链接为：https://github.com/LiuSH0402/EnzyProp
注意事项：
- 本数据集收录的文件均为训练完成的EnzyProp模型检查点；
- 本数据集未包含ESM2、ProtBert等基于HuggingFace的基础模型；
- hf_cache目录会在模型下载与使用过程中于本地生成。