Implementation of some similarity coefficients in conjunction with multiple UPGMA and neighbor-joining algorithms for enhancing phylogenetic trees.

Random Amplified Polymorphic DNA (RAPD) markers was used to analyze the genetic structure of five Indigenous Egyptian's chicken populations including Fayoumi, Dokki-4, Golden Montazah, Silver Montazah, and El- Salam, based on the taxa generated by the analysis of ten RAPD markers. The population genetic distances were estimated by using two cluster algorithms (UPGMA & NJ neighbor-joining) accompanied with ten similarity coefficients comprising Jaccard, Sorensen-Dice, Russel& Rao, Rogers & Tanimoto, Simple Matching, Pearson Phi, Lance &Williams, Mountford, Michael, and Kulchenzky-1. The results demonstrated that for almost all methodologies, the Jaccard and Sorensen-Dice followed by Simple Matching coefficients revealed extremely close results, because both of them exclude negative co-occurrences. Due to the fact that there is no guarantee that the DNA regions with negative co-occurrences between two strains are indeed identical, the use of coefficients such as Jaccard and Sorensen-Dice that do not include negative co- occurrences was imperative for closely related organisms along with the NJ neighbor-joining cluster algorithm.