Online tools for protein stability change prediction

Hover column header to see the description.

Abbreviations

ML – machine learning method
SVM – support vector machine
MLR – multiple linear regression
RF – random forest
NN – neural network
DBN – deep belief network
meta-predictor – consolidates several predictors in a consensus prediction
ML/NN facilitated – ML or NN was used, for example, to derive weights for energy function
– reported to suffer from overfitting (Fang, 2020).

Predictor Input Algorithm Multiple-point mutations Anti-symmetry Setting T Setting pH Year
I-Mutant 2 (doi, server) 1D ML (SVM) 2005
Eris (doi, server) 3D energy function (knowledge-based) 2007
I-Mutant 3 (doi, server) 1D ML (SVM) 2008
AUTO-MUTE (doi, server) 3D energy function (knowledge-based), ML facilitated (RF, SVM
REPTree, SVM Regression)
2008
iSTABLE (doi, server) 1D ML (SVM), meta-predictor 2013
NeEMO (doi, server) 3D NN 2014
INPS (doi, server) 1D ML (SVM) 2015
EASE-MM (doi, server) 1D ML (SVM) 2016
MAESTRO (doi, server) 3D energy function (knowledge-based), ML facilitated (MLR, NN, SVM) 2016
DNpro (paper, server) 1D NN (DBN) 2016
DynaMut (doi, server) 3D ML (RF), meta-predictor 2018
HoTMuSiC (doi, server) 3D energy function (knowledge-based), NN facilitated 2019
DeepDDG (doi, server) 3D NN
meta-predictor available
2019
BoostDDG (doi, server) 1D ML (XGBoost) 2020
PremPS (doi, server) 3D ML (RF regression) 2020
SAAFEC-SEQ (doi, server) 1D ML (gradient boosting decision tree) 2021


Другие материалы по теме: