Not getting in too deep: A practical deep learning approach to routine crystallisation image classification

Jamie Milne, Chen Qian, David Hargreaves, Julie C. Wilson*, Yinhai Wang

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

Abstract

Using a relatively small training set of ~16 thousand images from macrmolecular crystallisation experiments, we compare classification results obtained with four of the most widely- used convolutional deep-learning network architectures that can be implemented without the need for extensive computational resources. We show that the classifiers have different strengths that can be combined to provide an ensemble classifier achieving a classification accuracy comparable to that obtained by a large consortium initiative. We use eight classes to effectively rank the experimental outcomes, thereby providing detailed information that can be used with routine crystallography experiments to automatically identify crystal forma- tion for drug discovery and pave the way for further exploration of the relationship between crystal formation and crystallisation conditions.
Original languageEnglish
Article number0282562
Number of pages16
JournalPLoS ONE
Volume18
Issue number3
DOIs
Publication statusPublished - 9 Mar 2023

Bibliographical note

© 2023 Milne et al.

Cite this