Neural Networks: A Comprehensive Foundation
Neural Networks: A Comprehensive Foundation
Self-Organizing Maps
From Gene Trees to Species Trees
SIAM Journal on Computing
Analysis and visualization of gene expression data using self-organizing maps
Neural Networks - New developments in self-organizing maps
| Hi-index | 0.01 |
The incongruence between the gene trees and species remains a challenge in molecular phylogenetics. In this work, we propose a novel microarray approach to resolve this problem based on our previously proposed phylogenomic mining method. In our microarray approach, we first selected 28 genes from a set of statistically significant housekeeping genes from the S. cerevisiae cell cycle time series microarray data. Then we employed the BLAST and synteny criteria to identify homologs and orthologys of the selected genes among the genomes of other species. Finally, we applied the phylogenomic mining method for the aligned genes to infer the species phylogeny. The phylogenetic mining method used the self-organizing map mining, hierarchical clustering and entropy measure to concatenate the phylogenomically informative genes to infer species phylogenies. Compared with the original gene concatenation approach, our method not only overcome the ad-hoc mechanism and prohibitive phylogenetic computing problem of the species inference for the large number of taxa but also first integrated the microarray techniques in the species phylogeny inference.