Research On Protein Cryo-EM Fitting Algorithm

The function of a protein is absolutely dependent on its three-dimensional structure.And protein structure is always a hot topic in bioinformatics research. But sometimes thestructure can not be observed directly by the experimental conditions. So the proteinstructure prediction is worth studying, and the experimental data can help to improve theprediction accuracy. Cryo-EM data obtained by electronic microscope has high accuracy,can be regarded as the reference of prediction results.The fitting problem of protein structure and Cryo-EM data can be attributed to thefeature extraction and matching steps. The protein structure can move to each position ofCryo-EM in3D space, rotation at any degree. And a fitting score with Cryo-EM data canbe calculated, which shows the cross-correlation between them. In order to get highprecision results, there is a large search space. So it is important to get a effective searchmethod, to find a set of optimal solutions. Tabu-Search is an effective search algorithm,has been widely used in many scientific research field. And it can help to improve theefficiency in the fitting problem. And the fiiting’s score can help the researcher to pick upsome results which is better than the other ones amang the prediction structures.Finally, the practice shows that the function for the prediction protein structure andCryo-EM data fitting, and the design and realization method described in this paper isrational and feasible. There is a practical application value to improve the efficiency ofprediction of protein structure.