Research Article

Identification of Hot Spots in Protein Structures Using Gaussian Network Model and Gaussian Naive Bayes

Table 2

List of the top 20 measures based on tenfold cross validations of Gaussian Naive Bayes when using modes with the highest frequency inputted into the feature vector, where , the distance cutoff in GNM varies from 6.0 to 8.0 with step size of 0.1, and the sliding window size (sw) for multiple high modes ranges from 1 to 21 with step size of 2.

TopCutoffswsenspepreacc measure

17.41010.29240.89920.10800.87490.1577
27.41110.30410.88730.10120.86390.1518
37.41310.32750.87360.09760.85180.1503
47.22010.42690.82070.09030.80490.1491
57.41210.30990.88090.09800.85810.1489
67.21910.41520.82390.08950.80750.1473
77.32010.41520.82290.08910.80660.1467
87.11110.29240.88700.09750.86320.1462
97.51510.34500.85920.09280.83860.1462
107.1930.39770.83120.08950.81380.1461
117.31010.26900.89920.10020.87400.1460
127.41510.34500.85850.09230.83790.1457
137.11310.31580.87270.09370.85040.1446
147.51410.32750.86630.09270.84470.1445
157.51610.35090.85290.09050.83280.1439
167.41410.32750.86530.09210.84380.1438
177.61510.33330.86220.09160.84100.1438
187.51010.26320.90000.09890.87450.1438
197.3910.24560.90900.10120.88240.1433
207.11410.32750.86410.09140.84260.1429