International Journal of Proteomics

Review Article

A Survey of Computational Intelligence Techniques in Protein Function Prediction

Table 9

Summary of computational intelligence for protein function prediction by using gene expression data.
ReferenceCI techniquesPerformanceDatasets
[128]Multilayer perceptronTP rate: up to 79.6%, FP rate: up to 97%DNA array expression data
[129]MRF with BayesianSensitivity: 87%PPI, genetic interactions, highly correlated gene expression network, protein complex data, and structural properties
[130]SVMAccuracy: 89.44Gene expression data
[131]Genetic programming Accuracy: 92.50–98.7%Gene expression data
[132]Majority voting genetic programming Accuracy: 81.82%Gene expression data
[133]Genetic programming Accuracy: 94.9–99.27%Gene expression data
[134]Genetic programming Accuracy: 95.24–100%Gene expression values and constant values
[135]Fuzzy nearest clusterTop N accuracy: 65.27%Gene expression data
[136]-meansAccuracy: 0.16–0.24 PPI and gene expression data
[137]HypergraphAccuracy: 97.95%Gene expression data
[138]Discriminative local subspaces with SVMAverage precision: 63% and score: 0.44Gene expression data