Page Banner

United States Department of Agriculture

Agricultural Research Service


item Lavine, Barry
item Davidson, Charles
item Vander Meer, Robert - Bob
item Lahav, Sigal
item Soroker, Victoria
item Hefetz, Abraham

Submitted to: Journal of Chemometrics
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 1/9/2003
Publication Date: 10/2/2003
Citation: Lavine, B.K., Davidson, C., Vander Meer, R.K., Lahav, S., Soroker, V., Hefetz, A. 2003. Genetic algorithms for deciphering the complex chemosensory code of social insects. Journal of Chemometrics. v. 66. p. 51-62.

Interpretive Summary: Mixtures of chemicals released by ants often cause changes in ant behavior. These complex chemical mixtures can be separated and the amount of each chemical measured. It is important to be able to compare the pattern of chemicals found under a variety of conditions. However, this complex data is extremely difficult to interpret, since the chemical components may only differ in their relative amounts. Scientists have developed analytical methods to help deal with the complexity of the data, but are always looking for improvements. Scientists at the Center for Medical, Agricultural and Veterinary Entomology, Gainesville, FL; Department of Chemistry, Clarkson University, Potsdam, NY; and the Department of Zoology, Tel-Aviv University, Ramat Aviv, Israel have successfully demonstrated a new method for the analysis of complex multi-component chemical data. The model data set was generated from the chemical analysis of glandular extracts of the desert ant. The pattern of chemicals from this gland causes ants to be either accepted or attacked by other ants. The chemical patterns are very complex and can be analyzed only using special methods. The authors used a new method that combines artificial intelligence and evolutionary concepts to visually separate the chemical patterns more precisely than previous methods. In this case, we found that chemicals from the queen ant had no influence on whether or not other ants were accepted or attacked. The method we demonstrated here for the analysis of complex chemical mixtures will find many uses in solving otherwise intractable problems in diverse areas of research.

Technical Abstract: Chemical communication among social insects is often studied with chromatographic methods. The data generated in such studies may be complex and require pattern recognition techniques for interpretation. We present the analysis of gas chromatographic profiles of hydrocarbon extracts obtained from the cuticle and postpharyngeal glands of 400 Cataglyphis niger worker ants using a genetic algorithm (GA) for pattern recognition analysis. The GA was used to identify the factors influencing colony odor. The pattern recognition GA identifies features (i.e., chromatographic peaks) whose principal component plots show clustering of the samples on the basis of class. Because the largest principal components capture the bulk of the variance in the data, the peaks chosen by the GA primarily convey information about differences between the classes in a data set. As it trains, the pattern recognition GA focuses on those classes and or samples that are difficult to classify by boosting their class and sample weights. Samples or classes that consistently classify correctly are not as heavily weighted as samples or classes that are difficult to classify. Over time, the algorithm learns its parameters in a manner similar to a neural network. The proposed algorithm integrates aspects of artificial intelligence and evolutionary computations to yield a "smart" one-pass procedure for feature selection and pattern recognition. Utilizing the pattern recognition GA, two specific questions were addressed in this study: 1) does the overall hydrocarbon profile of the colony change with time? (2) Does the queen influence the hydrocarbon pattern of the colony?

Last Modified: 10/15/2017
Footer Content Back to Top of Page