基于scaffold的分子生成模型 + 筛选模型，对比它们

一、基于scaffold的分子生成模型

1、Paper "3D-SMGE:A Pipeline for Molecules Generate and Evaluate based on Scaffolds"

code:GitHub - ZheLi-Lab-Collaboration/3D-SMGE: A Pipeline for Molecules Generate and Evaluate based on Scaffolds

conda activate autogrow

下载模型然后放在主文件夹：

https://drive.google.com/file/d/14c65EiA8gWlGhIxEyHtlp-NL68qDqN7P/view?usp=drive_link

基于SMIELS + scaffold的分子生成

python SMG_3D.py generate 3D_SMG ./model/ 100 --scaffold 'CC(C1=CC=C(OC)C(OC)=C1)=O' --genMode mode1 --inputFormat smiles --chunk_size 100 --cuda --max_length 60 --file_name scaffold

ADMET Predicition

Firstly, you are supposed to move the generated molecules agg_smi.smi to the ./data folder.

python ./property_Pred/ADMET/general_admet/admet-pred.py --smi_path ../data/agg_smi.smi --csv_path ../data/smi

2、Learning to Extend Molecular Scaffolds with Structural Motifs

GitHub - microsoft/molecule-generation: Implementation of MoLeR: a generative model of molecular graphs which supports scaffold-constrained generation

二、筛选模型

【精选】分子性质预测预训练模型 + code使用方法-CSDN博客