OATML-Markslab/ProteinGym_v1

ProteinGym是一个用于评估蛋白质适应性预测和设计模型的基准测试套件，包含近300万个不同的突变。它分为四个独立的基准，基于预测目标和评估的突变类型。DMS_substitutions包含来自DMS实验的替换突变蛋白质，DMS_indels包含DMS实验中的插入-删除突变蛋白质。clinical_substitutions包含ClinVar数据库中标记为致病或良性的替换突变，clinical_indels包含ClinVar数据库中标记为致病的插入-删除突变以及GnomAD数据库中的常见良性突变。

ProteinGym is a benchmark suite for evaluating protein fitness prediction and design models, including nearly 3 million different mutations. It is split into four separate benchmarks based on the prediction target and the type of mutation assessed. DMS_substitutions includes proteins from DMS experiments measuring substitution mutations, DMS_indels includes proteins from DMS experiments measuring insertion-deletion (indel) mutations. clinical_substitutions includes substitution mutations from the ClinVar database labeled as pathogenic or benign, and clinical_indels includes a mix of pathogenic-labeled indel mutations from ClinVar and frequently occurring mutations from the GnomAD database serving as benign examples.