Skip to main content

Table 5 Performance comparison of different approaches using five-fold cross-validation for the benchmark data sets.

From: Predicting RNA-binding sites of proteins using support vector machines and evolutionary information

Data set Method Spec. (%) Sens. (%) Acc (%) MCC Threshold
RBP86 Jeong 2006 91.04 43.4 80.2 0.39 (0.41)* --
  PPRint 89.55 53.05 81.16 0.45 --
  RNAProB § 90.36 79.95 87.99 0.68 0.36
  RNAProB # 90.01 79.64 87.65 0.67 0.36
RBP109 RNABindR 93.00 38.00 84.80 0.35 --
  RNAProB § 93.88 64.62 89.70 0.58 0.35
  RNAProB # 94.14 60.63 89.36 0.56 0.35
RBP107 BindN-PCP& 69.84 66.28 69.32 0.27 --
  BindN-ALL& 75.70 65.78 74.25 -- --
  PPRint 75.54 70.09 75.43 0.32 --
  RNAProB § 80.87 77.14 80.44 0.42 0.11
  RNAProB # 80.65 73.62 79.84 0.40 0.12
  1. § presents the performance by five-fold cross-validation.
  2. # denotes the performance by a three-way data split procedure.
  3. * indicates the performance of weighted profiles by Jeong and Miyano [7].
  4. &BindN-PCP represents the results based only on physicochemical properties, while BindN-ALL shows the performance using physicochemical properties, relative solvent accessible surface area, and BLAST results.