Neural‐Symbolic Machine Learning for Retrosynthesis and Reaction Prediction
Problem
Rule-based expert systems for reaction prediction and retrosynthesis often fail because they ignore molecular context, causing reactivity conflicts.
Why it matters
Reliable computational synthesis planning could greatly accelerate chemical and pharmaceutical development.
What the AI did
Deep neural networks trained on 3.5 million reactions learn to resolve reactivity conflicts and prioritize suitable transformation rules for retrosynthesis and reaction prediction.
Results
Top-10 accuracy of 95% for retrosynthesis and 97% for reaction prediction on a validation set of almost 1 million reactions.
Impact
High benchmark accuracy on a large validation set, but no field deployment or independent external verification is stated.
Record
| Venue | Chemistry - A European Journal |
|---|---|
| Published | 2017 |
| Work type | article |
| Authors | Marwin Segler, Mark P. Waller |
| Citations | 549 (OpenAlex, at import time) |
| References | 32 |
| Model or method | deep neural networks (architecture not further specified) |
| Organization | not stated in source |
| Primary source | https://doi.org/10.1002/chem.201605499 |
| Open access | closed |
| Keywords | retrosynthesis, reaction prediction, deep neural networks, reactivity conflicts, chemistry |
Importance rating
significant — thousands to millions affected
Abstract calls reaction prediction and retrosynthesis cornerstones of organic chemistry; reliable computational synthesis planning serves a foundational field and pharma development, validated on ~1M reactions.