EMD: An Ensemble Algorithm for Discovering Regulatory Motifs in DNA Sequences
Author Information
Author(s): Hu Jianjun, Yang Yifeng, Kihara Daisuke
Primary Institution: Purdue University
Hypothesis
Can an ensemble approach improve the prediction accuracy of DNA regulatory motifs?
Conclusion
The EMD algorithm significantly improves the prediction accuracy of DNA regulatory motifs compared to stand-alone algorithms.
Supporting Evidence
- The EMD algorithm achieved a 22.4% improvement in nucleotide level prediction accuracy over the best stand-alone algorithm.
- All EMD algorithms outperformed the best single component algorithm in terms of nucleotide level accuracy.
- The ensemble approach is effective for improving both sensitivity and specificity in motif discovery.
Takeaway
The EMD algorithm combines results from multiple DNA motif discovery programs to find patterns in DNA more accurately, like how a team of friends can solve a puzzle better than one person alone.
Methodology
The EMD algorithm combines predictions from multiple motif discovery algorithms and runs them multiple times to improve accuracy.
Potential Biases
The algorithm may be biased if the component algorithms produce similar predictions.
Limitations
The performance of the EMD algorithm decreases for longer input sequences.
Statistical Information
P-Value
0.05
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website