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 PPfold(seed) - scored higher in this pairwise comparison

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

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


Overview

Metric PPfold(seed) CMfinder(seed)
MCC 0.600 > 0.585
Average MCC ± 95% Confidence Intervals 0.589 ± 0.116 > 0.578 ± 0.165
Sensitivity 0.455 > 0.440
Positive Predictive Value 0.798 > 0.786
Total TP 91 > 88
Total TN 20199 < 20201
Total FP 26 < 27
Total FP CONTRA 4 > 1
Total FP INCONS 19 < 23
Total FP COMP 3 = 3
Total FN 109 < 112
P-value 1.60044567142e-05

^top




Performance plots


  1. Comparison of performance of PPfold(seed) and CMfinder(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 PPfold(seed) and CMfinder(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 PPfold(seed) and CMfinder(seed)).

  3. Comparison of average Matthews Correlation Coefficients (MCCs) for PPfold(seed) and CMfinder(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 PPfold(seed) and CMfinder(seed)).

^top





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

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

Total Base Pair Counts
Total TP 91
Total TN 20199
Total FP 26
Total FP CONTRA 4
Total FP INCONS 19
Total FP COMP 3
Total FN 109
Total Scores
MCC 0.600
Average MCC ± 95% Confidence Intervals 0.589 ± 0.116
Sensitivity 0.455
Positive Predictive Value 0.798
Nr of predictions 13

^top



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

RNA Chain Rfam family MCC SENS PPV TP TN FP FP CONTRA FP INCONS FP COMP FN
RFA_00654 0.23 0.17 0.33 3 2406 6 0 6 0 15
RFA_00658 0.38 0.14 1.00 2 1126 1 0 0 1 12
RFA_00664 0.38 0.14 1.00 2 988 1 0 0 1 12
RFA_00708 0.38 0.14 1.00 2 1033 1 0 0 1 12
RFA_00749 0.78 0.75 0.82 9 892 2 0 2 0 3
RFA_00764 0.78 0.75 0.82 9 892 2 0 2 0 3
RFA_00765 0.78 0.75 0.82 9 892 2 0 2 0 3
RFA_00769 0.68 0.56 0.83 10 1941 2 1 1 0 8
RFA_00770 0.68 0.56 0.83 10 2004 2 1 1 0 8
RFA_00773 0.68 0.56 0.83 10 1941 2 1 1 0 8
RFA_00779 0.68 0.56 0.83 10 1941 2 1 1 0 8
RFA_00808 0.75 0.56 1.00 9 2007 0 0 0 0 7
RFA_00809 0.50 0.38 0.67 6 2136 3 0 3 0 10

^top



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

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

Total Base Pair Counts
Total TP 88
Total TN 20201
Total FP 27
Total FP CONTRA 1
Total FP INCONS 23
Total FP COMP 3
Total FN 112
Total Scores
MCC 0.585
Average MCC ± 95% Confidence Intervals 0.578 ± 0.165
Sensitivity 0.440
Positive Predictive Value 0.786
Nr of predictions 13

^top



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

RNA Chain Rfam family MCC SENS PPV TP TN FP FP CONTRA FP INCONS FP COMP FN
RFA_00654 0.38 0.22 0.67 4 2409 2 0 2 0 14
RFA_00658 0.40 0.29 0.57 4 1121 4 0 3 1 10
RFA_00664 0.40 0.29 0.57 4 983 4 0 3 1 10
RFA_00708 0.40 0.29 0.57 4 1028 4 0 3 1 10
RFA_00749 0.91 0.83 1.00 10 893 0 0 0 0 2
RFA_00764 0.91 0.83 1.00 10 893 0 0 0 0 2
RFA_00765 0.91 0.83 1.00 10 893 0 0 0 0 2
RFA_00769 0.71 0.50 1.00 9 1944 0 0 0 0 9
RFA_00770 0.71 0.50 1.00 9 2007 0 0 0 0 9
RFA_00773 0.00 0.00 0.00 0 1949 4 0 4 0 18
RFA_00779 0.71 0.50 1.00 9 1944 0 0 0 0 9
RFA_00808 0.68 0.56 0.82 9 2005 2 0 2 0 7
RFA_00809 0.41 0.38 0.46 6 2132 7 1 6 0 10

^top


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