Technology Category
Machine Learning: Multi-class/Multi-label Learning & Extreme ClassificationKnowledge Representation and Reasoning: Diagnosis and Abductive ReasoningReasoning under Uncertainty: Graphical Models
From Classification to Generation: An Open-Ended Paradigm for Adverse Drug Reaction Prediction Based on Graph-Motif Feature Fusion
🤖 AI Summary
This work addresses key challenges in adverse drug reaction prediction—namely data scarcity, closed-label assumptions, and insufficient modeling of label dependencies—by proposing a dual-graph representation architecture that fuses graph- and fragment-based features. It dynamically extracts multi-granular molecular fragments using BRICS and extended fragmentation rules. Innovatively, the multi-label classification task is reformulated as an autoregressive generation problem via a Transformer decoder, where positional embeddings explicitly capture dependency and co-occurrence patterns among over 10,000 labels. Retrosynthetic fragment analysis further enhances model interpretability. The method achieves an average performance gain of 20% (up to 38%) while expanding the label space from 200 to over 10,000 classes, substantially improving generalization and practical applicability.
Technology Category
Application Category
📝 Abstract
Computational biology offers immense potential for reducing the high costs and protracted cycles of new drug development through adverse drug reaction (ADR) prediction. However, current methods remain impeded by drug data scarcity-induced cold-start challenge, closed label sets, and inadequate modeling of label dependencies. Here we propose an open-ended ADR prediction paradigm based on Graph-Motif feature fusion and Multi-Label Generation (GM-MLG). Leveraging molecular structure as an intrinsic and inherent feature, GM-MLG constructs a dual-graph representation architecture spanning the atomic level, the local molecular level (utilizing fine-grained motifs dynamically extracted via the BRICS algorithm combined with additional fragmentation rules), and the global molecular level. Uniquely, GM-MLG pioneers transforming ADR prediction from multi-label classification into Transformer Decoder-based multi-label generation. By treating ADR labels as discrete token sequences, it employs positional embeddings to explicitly capture dependencies and co-occurrence relationships within large-scale label spaces, generating predictions via autoregressive decoding to dynamically expand the prediction space. Experiments demonstrate GM-MLG achieves up to 38% improvement and an average gain of 20%, expanding the prediction space from 200 to over 10,000 types. Furthermore, it elucidates non-linear structure-activity relationships between ADRs and motifs via retrosynthetic motif analysis, providing interpretable and innovative support for systematic risk reduction in drug safety.
Problem
Research questions and friction points this paper is trying to address.
adverse drug reaction prediction
cold-start challenge
closed label sets
label dependencies
drug safety
Innovation
Methods, ideas, or system contributions that make the work stand out.
Graph-Motif Feature Fusion
Multi-Label Generation
Transformer Decoder
Open-Ended ADR Prediction
Retrosynthetic Motif Analysis
🔎 Similar Papers
No similar papers found.
