Abstract:In the field of chemical synthesis planning, the accurate recommendation of reaction conditions is essential for achieving successful outcomes. This work introduces an innovative deep learning approach designed to address the complex task of predicting appropriate reagents, solvents, and reaction temperatures for chemical reactions. Our proposed methodology combines a multi-label classification model with a ranking model to offer tailored reaction condition recommendations based on relevance scores derived from anticipated product yields. To tackle the challenge of limited data for unfavorable reaction contexts, we employed the technique of hard negative sampling to generate reaction conditions that might be mistakenly classified as suitable, forcing the model to refine its decision boundaries, especially in challenging cases. Our developed model excels in proposing conditions where an exact match to the recorded solvents and reagents is found within the top-10 predictions 73% of the time. It also predicts temperatures within ± 20 of the recorded temperature in 89% of test cases. Notably, the model demonstrates its capacity to recommend multiple viable reaction conditions, with accuracy varying based on the availability of condition records associated with each reaction. What sets this model apart is its ability to suggest alternative reaction conditions beyond the constraints of the dataset. This underscores its potential to inspire innovative approaches in chemical research, presenting a compelling opportunity for advancing chemical synthesis planning and elevating the field of reaction engineering. Scientific contribution : The combination of multi-label classification and ranking models provides tailored recommendations for reaction conditions based on the reaction yields. A novel approach is presented to address the issue of data scarcity in negative reaction conditions through data augmentation.

Text-Augmented Multimodal LLMs for Chemical Reaction Condition Recommendation

Uncertainty-calibrated deep learning for rapid identification of reaction mechanisms

Chemist-X: Large Language Model-empowered Agent for Reaction Condition Recommendation in Chemical Synthesis

Automated electrosynthesis reaction mining with multimodal large language models (MLLMs)

ReLM: Leveraging Language Models for Enhanced Chemical Reaction Prediction

ChemVLM: Exploring the Power of Multimodal Large Language Models in Chemistry Area

ChemEval: A Comprehensive Multi-Level Chemical Evaluation for Large Language Models

Predictive Chemistry Augmented with Text Retrieval

Natural Language-Assisted Multi-modal Medication Recommendation

LMM Chemical Research with Document Retrieval

Enhancing chemical synthesis: a two-stage deep neural network for predicting feasible reaction conditions

A Self-feedback Knowledge Elicitation Approach for Chemical Reaction Predictions

ChemDFM-X: Towards Large Multimodal Model for Chemistry

ReacLLaMA: Merging chemical and textual information in chemical reactivity AI models

ReactXT: Understanding Molecular "Reaction-ship" via Reaction-Contextualized Molecule-Text Pretraining

Fine-tuning Large Language Models for Chemical Text Mining

Integrating Machine Learning and Large Language Models to Advance Exploration of Electrochemical Reactions

Generic Interpretable Reaction Condition Predictions with Open Reaction Condition Datasets and Unsupervised Learning of Reaction Center

Unified Deep Learning Model for Multitask Reaction Predictions with Explanation

Integrating Machine Learning and Large Language Models to Advance Wu Exploration of Electrochemical Reactions