Skip to main content
. 2016 Jun 11;32(12):i341–i350. doi: 10.1093/bioinformatics/btw280

Table 1.

Summary of DFLpred’s predictive model

Feature name Description rpb Coeff
WIN_IUP_fractionD0 Number of residues predicted with IUPred_struct not to be in structured regions in a sliding window divided by the window length. −1.00 −1.10
WIN_IUP_stdL Standard deviation of propensity scores from IUPred _long for residues in the sliding window. 0.70 6.60
WIN_AAind_avgAURR980118 Average value of AURR980118 AA index for all residues in the sliding window. −0.66 −4.58
WIN_AAind_avgPALJ810114 Average value of AURR980118 AA index for all residues in the sliding window. 0.57 3.32
Intercept of the linear function N/A −0.92

rpb: Normalized point-biserial correlation coefficient of a given feature with the annotation of DFLs in the training dataset; Coeff: coefficient of a given feature in the linear model generated with logistic regression using training dataset.