Genome-wide EST data mining approaches to resolving incongruence of molecular phylogenies.

Thirty-six single genes of 6 plants inferred 18 unique trees using maximum parsimony. Such incongruence is an important challenge. How to reconstruct the congruent tree is still one of the most challenges in molecular phylogenetics. For resolving this problem, a genome-wide EST data mining approach was systematically investigated by retrieving a large size of EST data of 144 shared genes of 6 green plants from GenBank. The results show that the concatenated alignments approach overcame incongruence among single-gene phylogenies and successfully reconstructed the congruent tree of 6 species with 100% jackknife support across each branch when 144 genes was used. Jackknife supports of correct branches increased with number of genes linearly, but the number of wrong branches also increased linearly. For inferring the congruent tree, a minimum of 30 genes were required. This approach may provide potential power in resolving conflictions of phylogenies.