Freezing firefly algorithm for efficient planted (ℓ, d) motif search.

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Motif discovery in biological sequences is crucial for understanding gene function, human diseases, and drug design.
• The (ℓ, d) motif search problem within Planted Motif Search (PMS) is a widely studied area.
• Current algorithms face challenges in efficiency and speed for complex motif detection tasks.

Purpose of the Study:

• To propose an efficient optimization algorithm, Freezing FireFly (FFF), for solving the (ℓ, d) motif search problem.
• To enhance the performance of the basic Firefly algorithm by incorporating local and global freezing strategies.
• To demonstrate the algorithm's effectiveness on simulated and real biological datasets.

Main Methods:

• Development of the Freezing FireFly (FFF) algorithm, integrating local and global freezing mechanisms.
• Application of FFF to simulated datasets for performance evaluation, specifically targeting the (50, 21) instance.
• Testing FFF on real-world datasets, including Chromatin Immunoprecipitation sequencing (ChIP-seq) data, and synthetic datasets.

Main Results:

• The FFF algorithm successfully resolved the (50, 21) motif search instance on a simulated dataset in just 1.47 minutes.
• FFF demonstrated significantly faster performance compared to state-of-the-art algorithms like Samselect, TraverStringRef, PMS8, qPMS9, AlignACE, FMGA, and GSGA.
• The algorithm showed superior speed on both real and synthetic biological sequence datasets.

Conclusions:

• The proposed Freezing FireFly (FFF) algorithm offers an efficient and effective solution for the (ℓ, d) motif search problem.
• FFF provides a substantial performance improvement over existing optimization algorithms in motif discovery.
• This advancement has implications for accelerating biological discoveries through enhanced motif detection.