Table 5 RCNOIK algorithm based on PPRA optimization

Step 1: Feature selection for residues in datasets;

Step 2: Calculate the sum of distance squares of residues and the average silhouette value to obtain the optimal k value;

Step 3: K-means is used to obtain the hot regions RCNOIK_RH1;

Step 4: \(RCNOIK\_RH1\) is optimized by PPRA to obtain hot region \(RCNOIK\_RH2\);

Step 5: \(RCNOIK\_RH2\) is optimized according the RCNO strategy;

Step 6: Input the optimized hot regions \(RCNOIK\_Best\_CH_i\).