De Novo Molecular Generation with Stacked Adversarial Model

TL;DR

This paper introduces a novel stacked adversarial autoencoder model for de novo drug molecule generation, improving validity and similarity to known drugs by learning primitive features and refining molecules in two stages.

Contribution

The paper presents a new stacked adversarial autoencoder architecture that enhances molecule validity and similarity, advancing generative models for drug discovery.

Findings

Generated molecules are more valid and similar to known drugs.

The model outperforms baseline methods on LINCS L1000 dataset.

Two-stage learning improves feature extraction and molecule refinement.

Abstract

Generating novel drug molecules with desired biological properties is a time consuming and complex task. Conditional generative adversarial models have recently been proposed as promising approaches for de novo drug design. In this paper, we propose a new generative model which extends an existing adversarial autoencoder (AAE) based model by stacking two models together. Our stacked approach generates more valid molecules, as well as molecules that are more similar to known drugs. We break down this challenging task into two sub-problems. A first stage model to learn primitive features from the molecules and gene expression data. A second stage model then takes these features to learn properties of the molecules and refine more valid molecules. Experiments and comparison to baseline methods on the LINCS L1000 dataset demonstrate that our proposed model has promising performance for molecular generation.