LoocasGoose/cpr_data

Conformal Protein Retrieval数据集包含了运行Conformal Protein Retrieval Gradio Space所需的大型数据文件。该数据集由预嵌入的UniProt蛋白质序列和SCOPE蛋白质域序列组成，还包括符合预测所需的FDR和FNR阈值文件，以及用于校准预测的概率文件。此外，还包括Protein-Vec模型的checkpoint和配置文件。

The Conformal Protein Retrieval dataset contains the large data files required to run the Conformal Protein Retrieval Gradio Space. This dataset consists of pre-embedded UniProt protein sequences and SCOPE protein domain sequences, as well as FDR and FNR threshold files for conformal prediction, and calibration probabilities for Venn-Abers prediction. Additionally, it includes the Protein-Vec model checkpoints and configuration files.