Dual-population Firefly Algorithm Based on Gender Differences for Detecting Protein Complexes.

To address the problem of high false positive/negative rate and low accuracy in protein complex detection, we propose the Dual-population Firefly Algorithm Based on Gender Differences (DFAGD) based on the unique core-attachment structure of protein complexes and the biological properties of fireflies. This method divides the detection of protein complexes into two phases. Firstly, a global search of the male population is used to detect the core proteins, and then a local search of the female population is used to detect the attachment proteins, which improves the accuracy of detection. In the male population strategy, the population diversity is redefined, and when the diversity falls below the threshold, a spring model is introduced to bring the population into the repulsion phase so that it does not fall into a local optimum. The female population selects elite and excellent individuals from the detection results of the male population to perform guided neighborhood searches, which can effectively improve detection accuracy. Finally, the effectiveness of the protein complex detection method is tested by comparing it to eight classical detection methods using four datasets of Saccharomyces cerevisiae proteins. [ABSTRACT FROM AUTHOR]