St A T I S T I C A L A N A L Y S I S O F T H E F A C T O R S I N F L U E N C I N G T H E P R O T E I N F O L D I N G R A T E

Proteins can spontaneously fold accurately and quickly from unstable denatured states to their unique stable nature states. Accurately predicting the folding rate is the primary key to understand the protein structure and its folding mechanism, and also is one of the important contents of modern molecular biology. And to find the core factors deciding the protein folding rate will ensure the prediction of protein folding rates rather accurately. There has been suggested that the main factors influencing the protein folding rate can be generalized three aspects: chain length, typological structure, amino acid composition and physicochemical properties. The preference of backbone torsion angle and the interaction of hydrophobic residues play an important role in the process of driving proteins folded from denatured states to their unique nature states. We put forward the concept of cumulative torsion angle indicated the preference of backbone torsion angle, and hydrophobic buried area indicated the interaction of hydrophobic residues, and analyzed their influence to protein folding rates.The results show: 1) both cumulative torsion angle and hydrophobic buried area are the factors determining the protein folding rates; 2) starting from the backbone torsion angles predicted by amino acid sequences, we can develop a model of predicting protein folding rates based on amino acid sequences; 3) besides the chain length, the information influencing the folding rates of backbone cumulative torsion angle mostly distributed in beta sheet and turn(or random coil) structure. 4) the cumulative torsion angle show strong structure topological properties; 5) the folding rates mostly controlled by effective cumulative torsion angle for ultrafast folding proteins.Using the three parameters, mean cumulative torsion angle, hydrophobic buried area and natural logarithm of absolute contact order, we established a multiple regression model to predict the protein folding rates. The correlation of calculated and experimental values of protein folding rates is 0.80 in the current protein dataset and shows better than other models.