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 RNASampler(seed) - scored higher in this pairwise comparison

  4. Performance of ProbKnot - scored lower in this pairwise comparison

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


Overview

Metric RNASampler(seed) ProbKnot
MCC 0.779 > 0.644
Average MCC ± 95% Confidence Intervals 0.774 ± 0.092 > 0.616 ± 0.162
Sensitivity 0.659 < 0.668
Positive Predictive Value 0.921 > 0.623
Total TP 442 < 448
Total TN 413762 > 413523
Total FP 105 < 457
Total FP CONTRA 9 < 80
Total FP INCONS 29 < 191
Total FP COMP 67 < 186
Total FN 229 > 223
P-value 2.22210847902e-08

^top




Performance plots


  1. Comparison of performance of RNASampler(seed) and ProbKnot. 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 RNASampler(seed) and ProbKnot).

  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 RNASampler(seed) and ProbKnot).

  3. Comparison of average Matthews Correlation Coefficients (MCCs) for RNASampler(seed) and ProbKnot. 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 RNASampler(seed) and ProbKnot).

^top





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

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

Total Base Pair Counts
Total TP 442
Total TN 413762
Total FP 105
Total FP CONTRA 9
Total FP INCONS 29
Total FP COMP 67
Total FN 229
Total Scores
MCC 0.779
Average MCC ± 95% Confidence Intervals 0.774 ± 0.092
Sensitivity 0.659
Positive Predictive Value 0.921
Nr of predictions 12

^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
RFA_00606 0.44 0.36 0.54 14 21295 18 7 5 6 25
RFA_00620 0.52 0.44 0.63 17 21918 15 2 8 5 22
RFA_00626 0.80 0.64 1.00 56 56560 5 0 0 5 31
RFA_00627 0.77 0.63 0.93 55 56894 8 0 4 4 32
RFA_00628 0.78 0.64 0.95 55 57233 12 0 3 9 31
RFA_00630 0.77 0.66 0.90 57 56890 15 0 6 9 30
RFA_00814 0.87 0.78 0.97 32 25167 6 0 1 5 9
RFA_00815 0.88 0.78 1.00 32 24499 8 0 0 8 9
RFA_00816 0.86 0.76 0.97 31 23188 6 0 1 5 10
RFA_00817 0.86 0.76 0.97 31 21913 1 0 1 0 10
RFA_00818 0.86 0.73 1.00 30 20271 6 0 0 6 11
RFA_00819 0.88 0.78 1.00 32 27934 5 0 0 5 9

^top



Performance of ProbKnot - scored lower in this pairwise comparison

1. Total counts & total scores for ProbKnot

Total Base Pair Counts
Total TP 448
Total TN 413523
Total FP 457
Total FP CONTRA 80
Total FP INCONS 191
Total FP COMP 186
Total FN 223
Total Scores
MCC 0.644
Average MCC ± 95% Confidence Intervals 0.616 ± 0.162
Sensitivity 0.668
Positive Predictive Value 0.623
Nr of predictions 12

^top



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

RNA Chain Rfam family MCC SENS PPV TP TN FP FP CONTRA FP INCONS FP COMP FN
RFA_00606 0.60 0.62 0.59 24 21280 33 7 10 16 15
RFA_00620 0.43 0.49 0.38 19 21895 44 8 23 13 20
RFA_00626 0.86 0.80 0.91 70 56539 32 1 6 25 17
RFA_00627 0.69 0.69 0.68 60 56865 47 6 22 19 27
RFA_00628 0.90 0.88 0.93 76 57209 31 1 5 25 10
RFA_00630 0.69 0.66 0.72 57 56874 41 4 18 19 30
RFA_00814 0.32 0.37 0.28 15 25146 45 8 31 6 26
RFA_00815 0.57 0.61 0.54 25 24485 35 5 16 14 16
RFA_00816 0.80 0.88 0.73 36 23171 20 9 4 7 5
RFA_00817 0.10 0.12 0.08 5 21883 57 18 39 0 36
RFA_00818 0.48 0.54 0.43 22 20250 38 13 16 9 19
RFA_00819 0.96 0.95 0.98 39 27926 34 0 1 33 2

^top


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