Retro-Rank-In: A Ranking-Based Approach for Inorganic Materials Synthesis Planning

2025-02-06

AI Summary

To address the low efficiency of traditional trial-and-error approaches and the poor generalizability of existing machine learning methods—largely constrained by reliance on known precursors—in inorganic materials inverse synthesis, this paper proposes the first bipartite graph-based pairwise ranking framework. It maps target materials and candidate precursors into a shared latent space and formulates inverse synthesis as a ranking task. The method integrates material representation embeddings, a dual-stream graph neural network, and pairwise ranking learning, trained on a deduplicated and debiased dataset. Compared to multi-label classification and other paradigms, our framework significantly enhances zero-shot generalization, successfully predicting experimentally validated novel reactions unseen during training (e.g., CrB + Al → Cr₂AlB₂). It establishes new state-of-the-art performance on out-of-distribution generalization and precursor ranking tasks.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Natural Language Processing: Code Generation / Program Synthesis from Natural Language
• Data Mining & Knowledge Management: Linked Open Data, Knowledge Graphs & KB Completion

Application Category

• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Retrosynthesis strategically plans the synthesis of a chemical target compound from simpler, readily available precursor compounds. This process is critical for synthesizing novel inorganic materials, yet traditional methods in inorganic chemistry continue to rely on trial-and-error experimentation. Emerging machine-learning approaches struggle to generalize to entirely new reactions due to their reliance on known precursors, as they frame retrosynthesis as a multi-label classification task. To address these limitations, we propose Retro-Rank-In, a novel framework that reformulates the retrosynthesis problem by embedding target and precursor materials into a shared latent space and learning a pairwise ranker on a bipartite graph of inorganic compounds. We evaluate Retro-Rank-In's generalizability on challenging retrosynthesis dataset splits designed to mitigate data duplicates and overlaps. For instance, for Cr2AlB2, it correctly predicts the verified precursor pair CrB + Al despite never seeing them in training, a capability absent in prior work. Extensive experiments show that Retro-Rank-In sets a new state-of-the-art, particularly in out-of-distribution generalization and candidate set ranking, offering a powerful tool for accelerating inorganic material synthesis.

Problem

• Addresses retrosynthesis planning for inorganic materials
• Improves generalization to new reactions via ranking
• Enhances prediction of precursor compounds accuracy

Innovation

• Ranking-based retrosynthesis approach
• Shared latent space embedding
• Bipartite graph learning