Related Experiment Videos
Subtle motifs: defining the limits of motif finding algorithms
1Department of Computer Science and Engineering, University of California San Diego, La Jolla, CA 92093, USA. keich@cs.ucsd.edu
Bioinformatics (Oxford, England)
|October 12, 2002
Summary
Subtle motifs are hard to distinguish from random patterns. This study defines the "motif twilight zone" where motif discovery algorithms struggle, offering tools to evaluate subtle motif detection.
Area of Science:
- Bioinformatics
- Computational Biology
- Genomics
Background:
- Subtle motifs are statistically indistinguishable from random motifs.
- Determining sequence length is crucial for motif discovery to avoid losing patterns or encountering random ones.
Purpose of the Study:
- Define the 'motif twilight zone' where algorithms face challenges with subtle motifs.
- Develop an objective tool for evaluating subtle motif finding algorithms.
- Assess the MULTIPROFILER algorithm's performance in detecting subtle motifs.
Main Methods:
- Statistical analysis to define the motif twilight zone.
- Development of a performance evaluation tool for motif finding algorithms.
- Application of the evaluation tool to the MULTIPROFILER algorithm.
Main Results:
- The motif twilight zone is defined, identifying conditions where random motifs rival true discoveries.
- An objective evaluation tool for subtle motif finding algorithms has been proposed.
- The MULTIPROFILER algorithm's success in detecting subtle motifs was evaluated using the new tools.
Conclusions:
- Understanding the motif twilight zone is critical for effective motif discovery.
- The proposed evaluation tool provides an objective measure for algorithm performance.
- The MULTIPROFILER algorithm demonstrates capability in subtle motif detection.