Predicting protein thermal stability changes upon point mutations using statistical potentials: Introducing HoTMuSiC

Abstract

The accurate prediction of the impact of an amino acid substitution on the thermal stability of a protein is a central issue in protein science, and is of key relevance for the rational optimization of various bioprocesses that use enzymes in unusual conditions. Here we present one of the first computational tools to predict the change in melting temperature ∆T m upon point mutations, given the protein structure and, when available, the melting temperature T m of the wild-type protein.The key ingredients of our model structure are standard and temperature-dependent statistical potentials, which are combined with the help of an artificial neural network. The model structure was chosen on the basis of a detailed thermodynamic analysis of the system. The parameters of the model were identified on a set of more than 1,600 mutations with experimentally measured ∆T m .The performance of our method was tested using a strict 5-fold cross-validation procedure, and was found to be significantly superior to that of competing methods. We obtained a root mean square deviation between predicted and experimental ∆T m values of 4.2°C that reduces to 2.9°C when ten percent outliers are removed. A webserver-based tool is freely available for non-commercial use at soft.dezyme.com. * fapucci@ulb.ac.be † rbourgeas@ulb.ac.be ‡ mrooman@ulb.ac.be 1 not peer-reviewed)

RiSLnet: Rapid identification of smart mutant libraries using protein structure network. Application to thermal stability enhancement

“…RiSLnet was able to significantly reduce the mutagenesis library when compared with the computational method, HoTMuSiC (Pucci, Bourgeas, & Rooman, 2016b). This method also takes a 3D structure of a protein as the input to predict ΔT m (T m mut -T m wt ) of 19 different substitutions for each residue.…”

Protein Thermal Stability Engineering Using Hotmusic

“…The ANN of T m -HoTMuSiC is composed of three layers, where the additional -hidden -layer gets activated by functions of the T m of the wild-type protein, and confer more weight to mesostable or thermostable perceptrons according to the thermal properties of the target protein (see Fig. 1b and [7] for details). A standard back-propagation algorithm was employed in the training of the neural network.…”

“…[14]. The results are reported in [7]. Here, in addition, we compared the HotMuSiC performances with popular ∆∆G prediction methods, namely PoPMuSiC [15], FoldX [16] and Rosetta [17].…”

“…This is primarily due to the huge size of the sequence space to be explored, and to the incomplete knowledge of the molecular mechanisms involved. Here we present some advances in the computational protein design field by describing the HoTMuSiC software [7] that we recently developed. HoTMuSiC is an efficient tool that, given the experimental or modelled three-dimensional (3D) structure of the target protein as input, screens the protein sequence and predicts the impact of all possible single amino acid substitutions on the protein melting temperature in just a few minutes.…”

Using machine learning to predict quantitative phenotypes from protein and nucleic acid sequences

“…Therefore, many current quantitative methods require a densely sampled sequence space, a highly specific biophysical model of the process being studied, or sacrifice specificity for sufficient examples necessary for optimizing model parameters. In particular these limitations are seen in the computational method used to describe a protein's thermodynamic properties, where highly accurate models of protein folding free energy require a threedimensional protein structure and a highly specific biophysical model (Yasuda et al, 2017;Pucci et al, 2016), extensive mutagenesis of the target protein family (Muk et al, 2019), or are purely trained with many protein families and therefore of limited specificity to any particular protein family (Montanucci et al, 2008;Li and Fang, 2010;Gromiha and Suresh, 2008;Li et al, 2019).…”