Skip to main content

Table 3 Comparison with other algorithms

From: Research on RNA secondary structure predicting via bidirectional recurrent neural network

Dataset Method SEN PPV ACC MCC
SPR VLDB 0.962 0.885 0.921 0.845
SVM 0.788 0.856 0.834 0.667
ProbKnot 0.793 0.744 0.772 0.546
LSTM 0.703 0.71 0.687 0.372
Cylofold * * * *
ASE VLDB 0.826 0.652 0.727 0.475
SVM 0.712 0.663 0.68 0.361
ProbKnot 0.734 0.564 0.613 0.247
LSTM 0.81 0.739 0.574 0.786
Cylofold 0.66 0.575 0.65 0.299
RFA VLDB 0.811 0.699 0.778 0.558
SVM 0.151 0.748 0.581 0.182
ProbKnot 0.793 0.555 0.648 0.339
LSTM 0.794 0.54 0.561 0.141
Cylofold 0.667 0.551 0.584 0.177
SRP VLDB 0.828 0.7 0.729 0.463
SVM 0.682 0.566 0.581 0.167
ProbKnot 0.807 0.598 0.638 0.300
LSTM 0.824 0.665 0.625 0.123
Cylofold 0.673 0.563 0.184 0.589
TMR VLDB 0.796 0.669 0.765 0.529
SVM 0.498 0.684 0.68 0.34
ProbKnot 0.635 0.388 0.533 0.109
LSTM 0.85 0.538 0.569 0.18
Cylofold 0.526 0.433 0.561 0.106
  1. The data in bold, italics, underline represent the optimal evaluation index values obtained by different algorithms on the same data set
  2. *Cylofold algorithm is unable to measure results on SPR data sets with many base deletion sequences