使用
模块信息
介绍文档
| 编号 | 405 | |
| 名称 | ADMET Prediction (v2) | |
| 封面 | ![]() | |
| 描述 | 基于机器学习的小分子ADMET性质预测模块,支持27种ADMET性质。 | |
| 标签 |
| |
| 作者 | WECOMPUT | |
| 发布时间 | 2024-11-28 00:00:00 | |
| 参考文献 | ||
目录导航
- 简介
- 参数说明
- Small Molecules
- Properties
- Predicted Results
- 结果说明
ADMET Prediction (v2)
简介
ADMET Prediction (v2)是一个基于机器学习的小分子ADMET性质预测模块。能快速批量预测小分子的ADMET性质,支持图注意力神经网络模型(GNN)、轻量梯度提升树模型(LGBM)、随机森林模型(RF)、梯度提升树模型(XGBT)4种常见高效的机器学习算法,分子特征支持分子指纹(Morgan FP)以及分子描述符(Descriptors)两种方法,能对小分子化合物库进行快速批量预测。模块支持27种ADMET性质,其中7种回归模型,20种分类模型。不同机器学习方法以及分子特征化方法预测性能如下:


模块自动选择最理想的机器学习算法和分子特征化方法的组合进行预测。
参数说明
Small Molecules
待预测的小分子文件,SDF格式。
Properties
ADMET预测列表,ADMET性质见结果说明部分。
Predicted Results
输出的预测结果文件,默认为predicted_results.csv
结果说明
输出结果中,如果是分类模型,输出0或1分类。如果是回归模型,预测出实际值。
ADMET性质信息如下:
| Dataset | Dataset Abbr. | ADMET Type | Dataset Type | Endpoints Description |
|---|---|---|---|---|
| Caco-2 (Cell Effective Permeability), Wang et al. | caco2 | Absorption | Regression | logPapp |
| PAMPA Permeability, NCATS | pampa | Absorption | Binary classification | high permeability (1) or low-to-moderate permeability (0) in PAMPA assay |
| HIA (Human Intestinal Absorption), Hou et al. | hia | Absorption | Binary classification | good permeability (1) or poor permeability (0) |
| Pgp (P-glycoprotein) Inhibition, Broccatelli et al. | pgp | Absorption | Binary classification | inhibitor (1) or non-inhibitor (0) |
| Bioavailability, Ma et al. | bioavailability | Absorption | Binary classification | High (1) or low (0) bioavailability |
| Lipophilicity, AstraZeneca | lipophilicity | Absorption | Regression | octanol/water distribution coefficient (logD at pH 7.4) |
| Solubility, AqSolDB | solubility | Absorption | Regression | logS |
| Hydration Free Energy, FreeSolv | freesolv | Absorption | Regression | Hydration Free Energy (kcal/mol) |
| BBB (Blood-Brain Barrier), Martins et al. | bbbp | Distribution | Binary classification | High (1) or low (0) blood-brain barrier penetration |
| PPBR (Plasma Protein Binding Rate), AstraZeneca | ppbr | Distribution | Regression | Plasma Protein Binding Rate (0-100) |
| CYP P450 2C19 Inhibition, Veith et al. | cyp2c19_inhibition | Metabolism | Binary Classification | P450 2C19 inhibitor (1) or non-inhibitor (0) |
| CYP P450 2D6 Inhibition, Veith et al. | cyp2d6_inhibition | Metabolism | Binary Classification | P450 2D6 inhibitor (1) or non-inhibitor (0) |
| CYP P450 3A4 Inhibition, Veith et al. | cyp3a4_inhibition | Metabolism | Binary Classification | P450 3A4 inhibitor (1) or non-inhibitor (0) |
| CYP P450 1A2 Inhibition, Veith et al. | cyp1a2_inhibition | Metabolism | Binary Classification | P450 1A2 inhibitor (1) or non-inhibitor (0) |
| CYP P450 2C9 Inhibition, Veith et al. | cyp2c9_inhibition | Metabolism | Binary Classification | P450 2C9 inhibitor (1) or non-inhibitor (0) |
| CYP2C9 Substrate, Carbon-Mangels et al. | cyp2c9_substrate | Metabolism | Binary Classification | CYP2C9 substrate (1) or non-substrate (0) |
| CYP2D6 Substrate, Carbon-Mangels et al. | cyp2d6_substrate | Metabolism | Binary Classification | CYP2CD6 substrate (1) or non-substrate(0) |
| CYP3A4 Substrate, Carbon-Mangels et al. | cyp3a4_substrate | Metabolism | Binary Classification | CYP3A4 substrate (1) or non-substrate(0) |
| Microsome Clearance, AstraZeneca | clearance_microsome | Excretion | Regression | Microsome Clearance (CL) |
| Acute Toxicity LD50 | ld50 | Toxicity | Regression | Acute Toxicity LD50 |
| hERG blockers | herg_blockers | Toxicity | Binary classification | hERG blockers (1) or non-blockers (0) |
| hERG Karim et al. | herg_karim | Toxicity | Binary classification | hERG blockers (1) or non-blockers (0) |
| Ames Mutagenicity | ames | Toxicity | Binary classification | high (1) or low (0) ames mutagenicity |
| DILI (Drug Induced Liver Injury) | dili | Toxicity | Binary classification | high (1) or low (0) drug induced liver injury |
| Skin Reaction | skin | Toxicity | Binary classification | high (1) or low (0) skin reaction |
| ClinTox | clintox | Toxicity | Binary classification | high (1) or low (0) ClinTox |
| Carcinogens | carcinogens | Toxicity | Binary classification | high (1) or low (0) Carcinogens |



