Related papers: MolFusion: Multimodal Fusion Learning for Molecular Representations via Multi-granularity Views

MolFusion: Multimodal Fusion Learning for Molecular Representations via Multi-granularity Views

URL: http://arxiv.org/abs/2406.18020v1
Date: Wed, 26 Jun 2024 02:26:50 GMT
Title: MolFusion: Multimodal Fusion Learning for Molecular Representations via Multi-granularity Views
Authors: Muzhen Cai, Sendong Zhao, Haochun Wang, Yanrui Du, Zewen Qiang, Bing Qin, Ting Liu,
Abstract summary: We propose a multi-granularity fusion method for combining molecular multi-modalities. MolFusion consists of two key components: MolSim, a molecular-level encoding component, and AtomAlign, an atomic-level encoding component. Results show that MolFusion effectively utilizes complementary multimodal information, leading to significant improvements in performance.
Score: 25.69424590542192
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Abstract: Artificial Intelligence predicts drug properties by encoding drug molecules, aiding in the rapid screening of candidates. Different molecular representations, such as SMILES and molecule graphs, contain complementary information for molecular encoding. Thus exploiting complementary information from different molecular representations is one of the research priorities in molecular encoding. Most existing methods for combining molecular multi-modalities only use molecular-level information, making it hard to encode intra-molecular alignment information between different modalities. To address this issue, we propose a multi-granularity fusion method that is MolFusion. The proposed MolFusion consists of two key components: (1) MolSim, a molecular-level encoding component that achieves molecular-level alignment between different molecular representations. and (2) AtomAlign, an atomic-level encoding component that achieves atomic-level alignment between different molecular representations. Experimental results show that MolFusion effectively utilizes complementary multimodal information, leading to significant improvements in performance across various classification and regression tasks.