Grading amino acid properties increased accuracies of single point mutation on protein stability prediction

We employed three predefined values, 0.1, 0.5, and 0.9 to represent 'weak', 'middle', and 'strong' groups for each amino acid property, and introduced two thresholds for each property to split twenty amino acids into one of the three groups according to their property values. Each amino acid can take only one out of three predefined values rather than twenty different values for each property. The complexity and noises in the encoding schemes were reduced in this way. More than 7% average accuracy improvement was found in the graded amino acid property encoding schemes by 20-fold cross validation. The overall accuracy of our method is more than 72% when performed on the independent test sets starting from sequence information with three-state prediction definitions.Grading numeric values of amino acid property can reduce the noises and complexity of input information. It is in accordance with biochemical concepts for amino acid properties and makes the input data simplified in the mean time. The idea of graded property encoding schemes may be applied to protein related predictions with machine learning approaches.Protein thermodynamic stability change upon single point mutations is a crucial problem that affects most protein engineering and molecular biology researches. Significant numbers of different prediction methods have been developed to predict the protein stability free energy change (ΔΔG) in last decades. While energy function-based approaches and statistical analysis were employed to compute the stability free energy change [1-14], machine learning approaches attracted more attention for increasing number of available experimental thermodynamic data in the ProTherm database [15-21]. Given the tertiary structure available, structure information based approaches generally performed better than sequence information based approaches in machine learning approaches [19]. The number of known protein structures, however, is less than one percent (0.45%) of the num