Structure-based prediction of BRAF mutation classes using machine-learning approaches.

Krebs, Fanny S; Britschgi, Christian; Pradervand, Sylvain; Achermann, Rita; Tsantoulis, Petros; Haefliger, Simon; Wicki, Andreas; Michielin, Olivier; Zoete, Vincent (2022). Structure-based prediction of BRAF mutation classes using machine-learning approaches. Scientific Reports, 12(1), p. 12528. Nature Publishing Group 10.1038/s41598-022-16556-x

Preview

Text s41598-022-16556-x.pdf - Published Version Available under License Creative Commons: Attribution (CC-BY). Download (1MB) | Preview

The BRAF kinase is attracting a lot of attention in oncology as alterations of its amino acid sequence can constitutively activate the MAP kinase signaling pathway, potentially contributing to the malignant transformation of the cell but at the same time rendering it sensitive to targeted therapy. Several pathologic BRAF variants were grouped in three different classes (I, II and III) based on their effects on the protein activity and pathway. Discerning the class of a BRAF mutation permits to adapt the treatment proposed to the patient. However, this information is lacking new and experimentally uncharacterized BRAF mutations detected in a patient biopsy. To overcome this issue, we developed a new in silico tool based on machine learning approaches to predict the potential class of a BRAF missense variant. As class I only involves missense mutations of Val600, we focused on the mutations of classes II and III, which are more diverse and challenging to predict. Using a logistic regression model and features including structural information, we were able to predict the classes of known mutations with an accuracy of 90%. This new and fast predictive tool will help oncologists to tackle potential pathogenic BRAF mutations and to propose the most appropriate treatment for their patients.

Interest & Impact

Downloads

34 since deposited on 27 Jul 2022
13 in the past 12 months

Citations

5 Citations in Web of Science ®
6 Citations in Scopus

Search

in Google Scholar™

Services

Actions (login required)

Edit item

Actions (login required)

Edit item