ProcedureT5: Adaptive Experimental Procedure Prediction with Data-Augmented Pre-Training and Multi-Source Data Integration

Key Points

• This research aims to develop a method that enhances the prediction of experimental procedures in chemical synthesis using data augmentation and multi-source integration.
• Utilized computer-aided synthesis planning techniques
• Implemented data-augmented pre-training
• Integrated data from multiple sources to improve prediction accuracy
• Demonstrated improved prediction of synthesis techniques
• Achieved faster synthesis route generation
• Highlighted increased effectiveness of data integration in prediction accuracy