Prediction of Drug-Kinase Binding Affinities with Focus on Conserved Protein Kinase Domain (CROSBI ID 727830)
Neobjavljeno sudjelovanje sa skupa | neobjavljeni prilog sa skupa | međunarodna recenzija
Podaci o odgovornosti
Oršolić, Davor ; Lučić, Bono ; Stepanić, Višnja ; Šmuc, Tomislav
engleski
Prediction of Drug-Kinase Binding Affinities with Focus on Conserved Protein Kinase Domain
Previous approaches implemented on drug-kinase binding affinity benchmark datasets show poor performance on rigorous test scenarios with unseen small compounds or protein kinase targets - thus limiting their real-world application. We represent methodology which relies on an ensemble approach, XGBoost trained on fingerprint based and protein kinase domain sequence-based similarity features - and graph convolutional networks (GCN) as more advanced representation learning predictive methodology. To assess the uncertainty surrounding model predictions, we defined a structure-based applicability domain with a focus on density of compound space in the training set.
kinase ; binding affinity
nije evidentirano
nije evidentirano
nije evidentirano
nije evidentirano
nije evidentirano
nije evidentirano
Podaci o prilogu
nije evidentirano
nije evidentirano
Podaci o skupu
ISMB/ECCB 2021 (29th Annual Conference)
poster
25.07.2021-30.07.2021
Virtual Event