logo
ResearchBunny Logo
Inferring experimental procedures from text-based representations of chemical reactions

Chemistry

Inferring experimental procedures from text-based representations of chemical reactions

A. C. Vaucher, P. Schwaller, et al.

This groundbreaking research by Alain C. Vaucher, Philippe Schwaller, Joppe Geluykens, Vishnu H. Nair, Anna Iuliano, and Teodoro Laino unveils advanced data-driven models capable of predicting synthesis steps from chemical equations. With an impressive dataset of 693,517 entries and innovative models like Transformer and BART, over 50% of predicted sequences require no human intervention for execution.

00:00
00:00
Playback language: English
Abstract
This work presents data-driven models for predicting the entire sequence of synthesis steps from a textual representation of a chemical equation, focusing on batch organic chemistry. A dataset of 693,517 chemical equations and associated action sequences was generated by processing experimental procedure text from patents using natural language models. Three models were trained: a nearest-neighbor model using reaction fingerprints, and two deep-learning sequence-to-sequence models (Transformer and BART). A chemist's analysis showed that predicted action sequences were adequate for execution without human intervention in over 50% of cases.
Publisher
Nature Communications
Published On
May 06, 2021
Authors
Alain C. Vaucher, Philippe Schwaller, Joppe Geluykens, Vishnu H. Nair, Anna Iuliano, Teodoro Laino
Tags
data-driven models
batch organic chemistry
chemical equations
prediction
natural language processing
sequence-to-sequence models

Related Publications

Explore these studies to deepen your understanding of the subject.

Text complexity and translation styles from the perspective of individuation: a case study of the English translations of *Pipa Xing*
Linguistics and Languages

Text complexity and translation styles from the perspective of individuation: a case study of the English translations of *Pipa Xing*

Y. Yu and C. Chang

Prediction of transition state structures of gas-phase chemical reactions via machine learning
Chemistry

Prediction of transition state structures of gas-phase chemical reactions via machine learning

S. Choi

Ideological representations of women in Jordanian folk proverbs from the perspective of cultural semiotics
Humanities

Ideological representations of women in Jordanian folk proverbs from the perspective of cultural semiotics

N. N. Al-khawaldeh, A. A. Banikalef, et al.

Machine vision-based detections of transparent chemical vessels toward the safe automation of material synthesis
Chemistry

Machine vision-based detections of transparent chemical vessels toward the safe automation of material synthesis

L. C. O. Tiong, H. J. Yoo, et al.

Listen, Learn & Level Up
Over 10,000 hours of research content in 25+ fields, available in 12+ languages.
No more digging through PDFs, just hit play and absorb the world's latest research in your language, on your time.
listen to research audio papers with researchbunny