Skip to main content

Table 4 FCRN and MLP performance on G Switch Proteins

From: Protein secondary structure prediction using a small training set (compact model) combined with a Complex-valued neural network approach

Method Observed j   Predicted j   Q j (%) Q 3 (%)
   H E C   
  H 680 0 42 94.19  
FCRN E 25 384 137 70.33 80.36
  C 51 20 61 46.22  
  H 691 0 31 95.71  
MLP E 38 394 114 72.17 79.22
  C 51 57 24 18.19  
  1. Both networks were trained with SSP 55