Data for: Deep Ranking Analysis by Power Eigenvectors (DRAPE): a polypharmacology case study

A dataset comprising 55 molecules described by seven criteria was used. The criteria are composed of binding activity values for each target expressed as half maximal activity concentration (AC50), based on the dose-response curves, thus the smaller the concentration, the more active the molecules. Seven targets are taken into account belonging to the nuclear receptors family: Estrogen Receptor Alpha (ERα), Estrogen Receptor Beta (ERβ), Farnesoid X Receptor (FXR), Progesterone Receptor (PR), Pregnane X Receptor (PXR), Peroxisome Proliferator-Activated Receptor Gamma (PPARγ) and Peroxisome Proliferator-Activated Receptor Delta (PPARδ). To create the dataset we collected from [22] the Tox21 databases [23, 24] of agonism/antagonism activity for the seven nuclear receptors.本研究采用一套涵盖55种分子的数据集，该数据集通过7项指标进行表征。这7项指标均基于剂量-反应曲线，以半数有效浓度（half maximal activity concentration, AC50）的形式表示各靶点的结合活性值，且浓度越低，对应分子的活性越强。本次涉及的7个靶点均属于核受体家族，分别为：雌激素受体α（Estrogen Receptor Alpha, ERα）、雌激素受体β（Estrogen Receptor Beta, ERβ）、法尼醇X受体（Farnesoid X Receptor, FXR）、孕激素受体（Progesterone Receptor, PR）、孕烷X受体（Pregnane X Receptor, PXR）、过氧化物酶体增殖物激活受体γ（Peroxisome Proliferator-Activated Receptor Gamma, PPARγ）以及过氧化物酶体增殖物激活受体δ（Peroxisome Proliferator-Activated Receptor Delta, PPARδ）。为构建该数据集，我们从文献[22]所收录的Tox21数据库[23,24]中获取了上述7种核受体的激动/拮抗活性相关数据。