Protein Structural Prediction based on hybridization of PSOFFA and CNN-BILSTM

Protein can be grouped in various classes like A (All ±), B (All
²
), C(±+
²
), and D(±/
²
). Numerous academics are working to increase the prediction accuracy, the fundamental difficulty in this case is correctly identifying homologous sequences and extracting the key properties. The demand for designing numerous sorts of pharmaceuticals is rising along with the prevalence of various diseases in human environments. Protein structure prediction play very important role in the same. In this paper, the selection of the feature derived by SCPRED was enhanced, leading to better classification. With the two datasets FC699 and 25PDB, we tested our method. In our study, we compare how well our method performs against other methods that use the same dataset. According to experimental findings, our proposed models’ accuracy, precision, specificity, sensitivity, Q3, and SOV values are superior than the most advanced prediction model.