همایش ملی بیوانفورماتیک ایران

صفحه اصلی / 4th international edition and 13th Iranian Conference on Bioinformatics

Accelerating Diffusion-Based Graph Generative Models for De Novo Drug Design via Hessian Trace Approximation

نویسندگان :

Negin Bagherpour¹ AmirHossein Heidari² Alireza Fotouhi Siahpirani³

1- دانشگاه تهران 2- دانشگاه تهران 3- دانشگاه تهران

کلمات کلیدی :

Diffusion-based models،Drug discovery،Policy optimization،Hessian approximation،Generative models

چکیده :

Diffusion-based models have recently gained prominence in various domains, including graph-structured data and de novo drug design, where they enable the generation of novel molecular structures and optimization of pharmaceutical candidates. In particular, methods such as “Digress” (Vignac et al. 2023) effectively incorporate diffusion processes to model complex interactions within graph data, while Graph Diffusion Policy Optimization (GDPO) (Liu et al. 2024) extends this idea by integrating policy optimization strategies to achieve higher-quality solutions. Compared to Digress, GDPO typically demonstrates higher convergence rate and more robust performance which makes it potentially better solution for advancing graph-based drug design problem and related applications. However, both Digress and GDPO rely heavily on gradient-based optimization, which is not as fast as necessary. On the other hand, computing the Hessian matrix directly to make use of second order methods and follow the curvature characteristics is computationally expensive. To address these issues, we introduce a novel idea to approximate the Hessian matrix with relatively low cost to better guide the optimization process, providing a more accurate and efficient estimation of curvature that leads to improved directions and lower number of iterations. We trained our model on the ZINC-250k dataset, a widely used collection of small molecules, and compared its performance with Digress and GDPO. Our approach demonstrates enhanced efficiency and superior performance over the existing diffusion-based generative models in de novo drug discovery.

لیست مقالات

لیست مقالات بایگانی شده

A New Method in Analyzing Drug-Drug Interaction Using Artificial Intelligence: Combination of Channel and Spatial Maps in Deep Attention Networks

Seyed Vahab Shojaedini - AmirReza Moeini

Investigating hsa-miR-146b and Its Targets TRAF6 and IRAK1 In Gastric Cancer Using Bioinformatic Tools

Ali Nosrat - Vahid Asghariazar - Elham Safarzadeh

Enhancing the Clustering and Sorting Procedures in the MAGUS Method for Multiple Sequence Alignment

Masih Hajsaeedi - Mohsen Hooshmand

Computational Analysis of the D92V Mutation in the EF-hand II Loop of the Mnemiopsin 2 Photoprotein

Hanieh Ramezani - Vahab Jafarian - Akram Shirdel - Khosrow Khalifeh - Fatemeh Khatami

Exploring Vaginal Microbiome Diversity in Early Pregnancy: Implications for Healthy Pregnancy and Miscarriage

Nasrin Vatankhah Hagigi - Mohammad Hossein Norouzi-Beirami

Attention based Graph Neural Network for Identifying Coding and Non-coding Breast Cancer Drivers

Bahar Mahdavi - Mitra Nemati Andavari - Mehdi Rajabizadeh - Mansoor Rezghi

Using interpretable deep learning models and multi-objective data for computational discovery of new drugs

Masoud Ahmadlou

Phylogenetic Insights into Enzybiotic with Novel Properties

Arman Hasannejad - Arad Ariaeenejad - Donya Afshar Jahanshahi - Mohammad Reza Zabihi - Shohreh Ariaeenejad - Kaveh Kavousi

Enzybiotic Activity in Marine Bacteria: Unveiling Cooperative and Competitive Dynamics for Antimicrobial Potential

Mohammad Reza Zabihi - Arad Ariaeenejad - Arman Hasannejad - Donya Afshar Jahanshahi - Shohreh Ariaeenejad - Kaveh Kavousi

BiogenExplorer: A Robust Solution for Rapid Gene Presence Detection and Allelic Diversity Analysis in large Genomic Datasets

Mohammadreza Najafi Disfani - Danial Ghofrani - Mahdiar Mansouri - Parastoo Saniee

ثمین همایش، سامانه مدیریت کنفرانس ها و جشنواره ها - نگارش 40.4.1