KIBA

该数据集名为KIBA，它将KIBA评分作为药物靶点亲和力的指标，这一评分综合了来自多种来源的激酶抑制剂生物活性数据。在亲和力值方面，范围从0.0到17.2，而未提供实验值的药物-靶点对则以“nan”值表示。该数据集涵盖了2111种独特药物，229个独特靶点，以及118,254个亲和力数据点，其任务是对药物-靶点结合亲和力进行预测。

This dataset is named KIBA, which employs the KIBA score as an indicator for drug-target binding affinity. This score integrates bioactivity data of kinase inhibitors collected from multiple sources. The affinity values range from 0.0 to 17.2, while drug-target pairs without corresponding experimental data are denoted by "nan". This dataset covers 2111 unique drugs, 229 unique targets, and 118,254 affinity data points, with its core task being the prediction of drug-target binding affinity.

搜集汇总

背景与挑战

背景概述

KIBA数据集是一个用于药物-靶标相互作用预测的综合数据集，包含117,657个相互作用对，涉及2,068种药物和229种蛋白质，通过整合多种生物活性类型来预测结合亲和力。