Skip to main content

Table 2 Performance on disulfide connectivity prediction obtained with different SVR-based methods

From: Prediction of disulfide connectivity in proteins with machine-learning methods and correlated mutations

# bonds SVR SVR+iCOV SVR+MI SVR+MI+iCOV
  Pb = Rb Qp Pb = Rb Qp Pb = Rb Qp Pb = Rb Qp
2 75 75 76 76 73 73 76 76
3 60 48 62.8 55.3 59.6 50.6 62.8 55.3
4 57 44 67.1 51.2 61 46.3 67.7 51.2
5 46 19 55.1 27 54.1 29.7 58.9 32.4
All 60 54 65.2 58.6 61.9 55.5 66.2 59.3
  1. # bonds: number of disulfide bonds; MIp: corrected Mutual Information; iCOV: sparse inverse COVariance estimation; SVR: Support Vector Regression; and their combinations as indicated. For details see Methods. Results are evaluated on the PDBCYS dataset [12]. SVR results are taken from [12]. For index definition see Performance measures.