Publication Date:
2014-10-01
Description:
Background: Studies of functional modules in a Protein-Protein Interaction (PPI) network contribute greatly to theunderstanding of biological mechanisms. With the development of computing science,computational approaches have played an important role in detecting functional modules. Results: We present a new approach using multi-agent evolution for detection of functional modules in PPInetworks. The proposed approach consists of two stages: the solution construction for agents in apopulation and the evolutionary process of computational agents in a lattice environment, where eachagent corresponds to a candidate solution to the detection problem of functional modules in a PPInetwork. First, the approach utilizes a connection-based encoding scheme to model an agent, andemploys a random-walk behavior merged topological characteristics with functional information toconstruct a solution. Next, it applies several evolutionary operators, i. e., competition, crossover, andmutation, to realize information exchange among agents as well as solution evolution. Systematicexperiments have been conducted on three benchmark testing sets of yeast networks. Experimentalresults show that the approach is more effective compared to several other existing algorithms. Conclusions: The algorithm has the characteristics of outstanding recall, F-measure, sensitivity and accuracy whilekeeping other competitive performances, so it can be applied to the biological study which requireshigh accuracy.
Electronic ISSN:
1471-2105
Topics:
Biology
,
Computer Science
Permalink