Skip to main content

Table 5 The performance of LPFS for disease gene selection and sample label prediction

From: Label propagation-based semi-supervised feature selection on decoding clinical phenotypes with RNA-seq data

Experiment Hamming loss One-error Coverage AUC AUPR
Q20 versus Q80 0.210 ± 0.027 0.676 ± 0.081 0.382 ± 0.404 0.513 ± 0.064 0.193 ± 0.024
Q20 versus Q92 0.220 ± 0.021 0.707 ± 0.063 0.397 ± 0.313 0.524 ± 0.060 0.211 ± 0.024
Q20 versus Q111 0.229 ± 0.024 0.733± 0.071 0.410 ± 0.353 0.556 ± 0.058 0.186 ± 0.020
Q20 versus Q140 0.226 ± 0.016 0.724 ± 0.048 0.406 ± 0.241 0.570 ± 0.066 0.210 ± 0.031
Q20 versus Q175 0.226 ± 0.015 0.726 ± 0.046 0.407 ± 0.232 0.605 ± 0.067 0.226 ± 0.015