Limits...
COFOLD: an RNA secondary structure prediction method that takes co-transcriptional folding into account.

Proctor JR, Meyer IM - Nucleic Acids Res. (2013)

Bottom Line: These aim to predict the most stable RNA structure.There exists by now ample experimental and theoretical evidence that the process of structure formation matters and that sequences in vivo fold while they are being transcribed.Here, we present a conceptually new method for predicting RNA secondary structure, called CoFold, that takes effects of co-transcriptional folding explicitly into account.

View Article: PubMed Central - PubMed

Affiliation: Centre for High-Throughput Biology, University of British Columbia, 2125 East Mall, Vancouver, BC, V6T 1Z4, Canada.

ABSTRACT
Existing state-of-the-art methods that take a single RNA sequence and predict the corresponding RNA secondary structure are thermodynamic methods. These aim to predict the most stable RNA structure. There exists by now ample experimental and theoretical evidence that the process of structure formation matters and that sequences in vivo fold while they are being transcribed. None of the thermodynamic methods, however, consider the process of structure formation. Here, we present a conceptually new method for predicting RNA secondary structure, called CoFold, that takes effects of co-transcriptional folding explicitly into account. Our method significantly improves the state-of-art in terms of prediction accuracy, especially for long sequences of >1000 nt in length.

Show MeSH
Changes in prediction accuracy for the structures predicted by CoFold for individual sequences. We report the prediction accuracy for base pairs of the long data set in terms of absolute changes by comparing the prediction accuracy of the structures predicted by CoFold with those predicted by RNAfold. The left plot shows change of the true positive rate () and PPV (). The right plot shows changes in true positive rate () and false positive rate (). TP denotes the numbers of true positives, TN the true negatives, FP the false positives and FN the false negatives.
© Copyright Policy - creative-commons
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC3643587&req=5

gkt174-F2: Changes in prediction accuracy for the structures predicted by CoFold for individual sequences. We report the prediction accuracy for base pairs of the long data set in terms of absolute changes by comparing the prediction accuracy of the structures predicted by CoFold with those predicted by RNAfold. The left plot shows change of the true positive rate () and PPV (). The right plot shows changes in true positive rate () and false positive rate (). TP denotes the numbers of true positives, TN the true negatives, FP the false positives and FN the false negatives.

Mentions: Compared with RNAfold, which is the state-of-the-art thermodynamic RNA structure prediction method, CoFold predicts 7% more known base pairs at 6% higher specificity than RNAfold, thereby increasing the MCC by 6% [MCC (RNAfold) = 42.81%, MCC (CoFold) = 49.10%] (Table 2). This improvement in overall performance accuracy can be attributed to a simultaneous increase of the positive predictive value (PPV) and the true positive rate (TPR) for almost all individual sequences (Figure 2 left) and a simultaneous slight decrease of the false positive rate (FPR) (Figure 2 right). Both RNAfold and CoFold use the default Turner 1999 free-energy parameters (5). Combining CoFold with the Andronescu 2007 free-energy parameters (6) (CoFold-A) increases the sensitivity and specificity by a further 4% [MCC (CoFold-A) = 53.70%]. Doing the same with RNAfold (RNAfold-A) also increases the sensitivity and specificity with respect to RNAfold, but it results in a smaller performance increase than for CoFold [MCC (RNAfold-A) = 48.17%, MCC (CoFold) = 49.10%]. Although CoFold only depends on two free parameters, the Andronescu 2007 free-energy model (6) comprises 363 free parameters that were trained using machine-learning techniques.Figure 2.


COFOLD: an RNA secondary structure prediction method that takes co-transcriptional folding into account.

Proctor JR, Meyer IM - Nucleic Acids Res. (2013)

Changes in prediction accuracy for the structures predicted by CoFold for individual sequences. We report the prediction accuracy for base pairs of the long data set in terms of absolute changes by comparing the prediction accuracy of the structures predicted by CoFold with those predicted by RNAfold. The left plot shows change of the true positive rate () and PPV (). The right plot shows changes in true positive rate () and false positive rate (). TP denotes the numbers of true positives, TN the true negatives, FP the false positives and FN the false negatives.
© Copyright Policy - creative-commons
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC3643587&req=5

gkt174-F2: Changes in prediction accuracy for the structures predicted by CoFold for individual sequences. We report the prediction accuracy for base pairs of the long data set in terms of absolute changes by comparing the prediction accuracy of the structures predicted by CoFold with those predicted by RNAfold. The left plot shows change of the true positive rate () and PPV (). The right plot shows changes in true positive rate () and false positive rate (). TP denotes the numbers of true positives, TN the true negatives, FP the false positives and FN the false negatives.
Mentions: Compared with RNAfold, which is the state-of-the-art thermodynamic RNA structure prediction method, CoFold predicts 7% more known base pairs at 6% higher specificity than RNAfold, thereby increasing the MCC by 6% [MCC (RNAfold) = 42.81%, MCC (CoFold) = 49.10%] (Table 2). This improvement in overall performance accuracy can be attributed to a simultaneous increase of the positive predictive value (PPV) and the true positive rate (TPR) for almost all individual sequences (Figure 2 left) and a simultaneous slight decrease of the false positive rate (FPR) (Figure 2 right). Both RNAfold and CoFold use the default Turner 1999 free-energy parameters (5). Combining CoFold with the Andronescu 2007 free-energy parameters (6) (CoFold-A) increases the sensitivity and specificity by a further 4% [MCC (CoFold-A) = 53.70%]. Doing the same with RNAfold (RNAfold-A) also increases the sensitivity and specificity with respect to RNAfold, but it results in a smaller performance increase than for CoFold [MCC (RNAfold-A) = 48.17%, MCC (CoFold) = 49.10%]. Although CoFold only depends on two free parameters, the Andronescu 2007 free-energy model (6) comprises 363 free parameters that were trained using machine-learning techniques.Figure 2.

Bottom Line: These aim to predict the most stable RNA structure.There exists by now ample experimental and theoretical evidence that the process of structure formation matters and that sequences in vivo fold while they are being transcribed.Here, we present a conceptually new method for predicting RNA secondary structure, called CoFold, that takes effects of co-transcriptional folding explicitly into account.

View Article: PubMed Central - PubMed

Affiliation: Centre for High-Throughput Biology, University of British Columbia, 2125 East Mall, Vancouver, BC, V6T 1Z4, Canada.

ABSTRACT
Existing state-of-the-art methods that take a single RNA sequence and predict the corresponding RNA secondary structure are thermodynamic methods. These aim to predict the most stable RNA structure. There exists by now ample experimental and theoretical evidence that the process of structure formation matters and that sequences in vivo fold while they are being transcribed. None of the thermodynamic methods, however, consider the process of structure formation. Here, we present a conceptually new method for predicting RNA secondary structure, called CoFold, that takes effects of co-transcriptional folding explicitly into account. Our method significantly improves the state-of-art in terms of prediction accuracy, especially for long sequences of >1000 nt in length.

Show MeSH