CompaRNA - on-line benchmarks of RNA structure prediction methods
Home

Methods
Datasets
Rankings
RNA 2D Atlas

Help
FAQ

Contact us
RSS feeds
Twitter

Table of contents:

  1. Overview

  2. Performance Plots

  3. Performance of Sfold - scored higher in this pairwise comparison

  4. Performance of RNASampler(seed) - scored lower in this pairwise comparison

  5. Compile and download dataset for Sfold & RNASampler(seed) [.zip] - may take several seconds...


Overview

Metric Sfold RNASampler(seed)
MCC 0.505 > 0.482
Average MCC ± 95% Confidence Intervals 0.535 ± 0.158 > 0.494 ± 0.152
Sensitivity 0.421 > 0.342
Positive Predictive Value 0.610 < 0.684
Total TP 197 > 160
Total TN 91729 < 91818
Total FP 188 > 107
Total FP CONTRA 13 > 8
Total FP INCONS 113 > 66
Total FP COMP 62 > 33
Total FN 271 < 308
P-value 2.78993384543e-08

^top




Performance plots


  1. Comparison of performance of Sfold and RNASampler(seed). Positive Predictive Value (PPV) is plotted against sensitivity. Each dot represents a single test of each method. See tables below for raw data (individual counts for Sfold and RNASampler(seed)).

  2. Average Matthews Correlation Coefficients (MCC) with 95% confidence intervals (CIs) were plotted for different RNA families, for which at least 3 members were present in the benchmarking dataset. 'n' denotes the number of MCCs used to calculate the average and CI. See tables below for raw data (individual counts for Sfold and RNASampler(seed)).

  3. Comparison of average Matthews Correlation Coefficients (MCCs) for Sfold and RNASampler(seed). The whiskers correspond to 95% confidence intervals (CIs). 'n' denotes the number of MCCs used to calculate average MCCs and CIs. See tables below for raw data (individual counts for Sfold and RNASampler(seed)).

^top





Performance of Sfold - scored higher in this pairwise comparison

1. Total counts & total scores for Sfold

Total Base Pair Counts
Total TP 197
Total TN 91729
Total FP 188
Total FP CONTRA 13
Total FP INCONS 113
Total FP COMP 62
Total FN 271
Total Scores
MCC 0.505
Average MCC ± 95% Confidence Intervals 0.535 ± 0.158
Sensitivity 0.421
Positive Predictive Value 0.610
Nr of predictions 14

^top



2. Individual counts for Sfold [ download as .csv ]

RNA Chain Rfam family MCC SENS PPV TP TN FP FP CONTRA FP INCONS FP COMP FN
2KDQ_B 0.95 0.91 1.00 10 396 0 0 0 0 1
2LC8_A 0.45 0.35 0.58 7 1528 5 0 5 0 13
3A3A_A 0.87 0.76 1.00 28 3627 0 0 0 0 9
3IVN_B 0.78 0.61 1.00 19 2327 0 0 0 0 12
3J3E_8 0.06 0.06 0.08 2 7477 32 2 22 8 31
3J3F_8 0.35 0.33 0.38 12 12214 34 3 17 14 24
3JYX_4 0.35 0.30 0.42 10 12222 23 3 11 9 23
3O58_3 0.43 0.34 0.55 12 12381 17 0 10 7 23
3RKF_A 0.76 0.59 1.00 20 2191 0 0 0 0 14
3SD1_A 0.65 0.50 0.84 21 3891 4 1 3 0 21
3W3S_B 0.57 0.53 0.64 21 4720 13 1 11 1 19
3ZEX_C 0.32 0.21 0.48 11 14173 26 1 11 14 41
4A1C_2 0.16 0.15 0.17 5 11751 34 2 23 9 28
4JF2_A 0.78 0.61 1.00 19 2831 0 0 0 0 12

^top



Performance of RNASampler(seed) - scored lower in this pairwise comparison

1. Total counts & total scores for RNASampler(seed)

Total Base Pair Counts
Total TP 160
Total TN 91818
Total FP 107
Total FP CONTRA 8
Total FP INCONS 66
Total FP COMP 33
Total FN 308
Total Scores
MCC 0.482
Average MCC ± 95% Confidence Intervals 0.494 ± 0.152
Sensitivity 0.342
Positive Predictive Value 0.684
Nr of predictions 14

^top



2. Individual counts for RNASampler(seed) [ download as .csv ]

RNA Chain Rfam family MCC SENS PPV TP TN FP FP CONTRA FP INCONS FP COMP FN
2KDQ_B 0.95 0.91 1.00 10 396 0 0 0 0 1
2LC8_A -0.01 0.00 0.00 0 1530 10 0 10 0 20
3A3A_A 0.70 0.49 1.00 18 3637 0 0 0 0 19
3IVN_B 0.78 0.61 1.00 19 2327 0 0 0 0 12
3J3E_8 0.12 0.09 0.17 3 7485 22 2 13 7 30
3J3F_8 0.44 0.33 0.57 12 12225 15 1 8 6 24
3JYX_4 0.39 0.30 0.50 10 12226 13 2 8 3 23
3O58_3 0.45 0.34 0.60 12 12383 12 2 6 4 23
3RKF_A 0.76 0.59 1.00 20 2191 0 0 0 0 14
3SD1_A 0.60 0.38 0.94 16 3899 1 0 1 0 26
3W3S_B 0.55 0.30 1.00 12 4741 1 0 0 1 28
3ZEX_C 0.32 0.19 0.53 10 14177 13 1 8 4 42
4A1C_2 0.31 0.24 0.40 8 11761 20 0 12 8 25
4JF2_A 0.57 0.32 1.00 10 2840 0 0 0 0 21

^top


Matthews Correlation Coeffient, Sensitivity and Positive Predictive Value have been calculated based on the paper by Gardener & Giegerich, 2004.