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 ContextFold - scored higher in this pairwise comparison

  4. Performance of RSpredict(20) - scored lower in this pairwise comparison

  5. Compile and download dataset for ContextFold & RSpredict(20) [.zip] - may take several seconds...


Overview

Metric ContextFold RSpredict(20)
MCC 0.759 > 0.672
Average MCC ± 95% Confidence Intervals 0.783 ± 0.209 > 0.671 ± 0.248
Sensitivity 0.669 > 0.549
Positive Predictive Value 0.864 > 0.830
Total TP 89 > 73
Total TN 14630 < 14645
Total FP 15 < 17
Total FP CONTRA 3 < 6
Total FP INCONS 11 > 9
Total FP COMP 1 < 2
Total FN 44 < 60
P-value 0.0

^top




Performance plots


  1. Comparison of performance of ContextFold and RSpredict(20). 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 ContextFold and RSpredict(20)).

  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 ContextFold and RSpredict(20)).

  3. Comparison of average Matthews Correlation Coefficients (MCCs) for ContextFold and RSpredict(20). 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 ContextFold and RSpredict(20)).

^top





Performance of ContextFold - scored higher in this pairwise comparison

1. Total counts & total scores for ContextFold

Total Base Pair Counts
Total TP 89
Total TN 14630
Total FP 15
Total FP CONTRA 3
Total FP INCONS 11
Total FP COMP 1
Total FN 44
Total Scores
MCC 0.759
Average MCC ± 95% Confidence Intervals 0.783 ± 0.209
Sensitivity 0.669
Positive Predictive Value 0.864
Nr of predictions 5

^top



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

RNA Chain Rfam family MCC SENS PPV TP TN FP FP CONTRA FP INCONS FP COMP FN
PDB_00571 0.85 0.76 0.95 19 3301 1 1 0 0 6
PDB_00828 0.86 0.74 1.00 20 2465 0 0 0 0 7
PDB_00829 0.86 0.75 1.00 18 2260 0 0 0 0 6
PDB_01020 0.86 0.74 1.00 17 2261 1 0 0 1 6
PDB_01073 0.48 0.44 0.54 15 4343 13 2 11 0 19

^top



Performance of RSpredict(20) - scored lower in this pairwise comparison

1. Total counts & total scores for RSpredict(20)

Total Base Pair Counts
Total TP 73
Total TN 14645
Total FP 17
Total FP CONTRA 6
Total FP INCONS 9
Total FP COMP 2
Total FN 60
Total Scores
MCC 0.672
Average MCC ± 95% Confidence Intervals 0.671 ± 0.248
Sensitivity 0.549
Positive Predictive Value 0.830
Nr of predictions 5

^top



2. Individual counts for RSpredict(20) [ download as .csv ]

RNA Chain Rfam family MCC SENS PPV TP TN FP FP CONTRA FP INCONS FP COMP FN
PDB_00571 0.66 0.56 0.78 14 3303 5 0 4 1 11
PDB_00828 0.82 0.74 0.91 20 2463 2 2 0 0 7
PDB_00829 0.70 0.63 0.79 15 2259 4 2 2 0 9
PDB_01020 0.84 0.78 0.90 18 2258 3 2 0 1 5
PDB_01073 0.34 0.18 0.67 6 4362 3 0 3 0 28

^top


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