# Precision fragment addition: domain-specific DeepFrag2 models for smarter lead optimization

## TL;DR

DeepFrag2 is an improved machine-learning tool that helps chemists optimize drug candidates by adding suitable molecular fragments.

## Contribution

The study introduces DeepFrag2, which enhances lead optimization accuracy by using domain-specific models trained on fragment properties and receptor classes.

## Key findings

- DeepFrag2 models trained on specific fragment properties improve prediction accuracy.
- Fine-tuning DeepFrag2 on specific receptor classes further boosts performance.
- DeepFrag2 is released as open-source software for free use.

## Abstract

This study introduces a series of machine-learning models based on DeepFrag, our previously published tool designed to guide small-molecule lead optimization through fragment addition. We demonstrate enhanced accuracy by training new DeepFrag models to predict optimizing fragments with specific sizes and chemical properties. Additionally, we show that DeepFrag accuracy improves when fine-tuned on specific receptor classes. These targeted models should prove valuable for medicinal chemists with predetermined insights into suitable molecular fragment characteristics (such as preferred size ranges, charge states, or aromaticity) or those conducting optimization campaigns against specific drug-target classes with many known ligands. To encourage adoption, we release DeepFrag2 under the open-source MIT license. Interested users can download DeepFrag2 free of charge without registration from https://durrantlab.com/deepfrag2/.

DeepFrag2, a machine-learning tool for lead optimization via fragment addition, is more accurate when trained on fragments with specific sizes, charge states, or aromaticity. Fine-tuning on specific receptor classes further boosts performance.