Artificial neural networks and genetic dissimilarity among saladette type dwarf tomato plant populations

Abstract

Studies have shown that dwarf plants have the potential for use in obtaining hybrids. The aim of this study was to evaluate the agronomic potential and genetic dissimilarity of saladette type dwarf tomato plant populations through the use of artificial neural networks (ANNs). The following traits were analyzed: mean fruit weight, transverse and longitudinal fruit diameter, fruit shape, pulp thickness, locule number, internode length, soluble solids content, and β-carotene, lycopene, and leaf zingiberene contents. A dendrogram obtained by the unweighted pair-group method with arithmetic mean (UPGMA) and Kohonen self-organizing maps (SOM) agreed in the distinction of the BC1F3 populations from the dwarf donor parent. SOM was more consistent in identifying the genetic similarities among the BC1F3 dwarf tomato plant populations and allowed for the determination of weights of each variable in the cluster formation. The UFU SDi 13-1 BC1F3 population was revealed to be a promising option for obtaining saladette type dwarf tomato plant lines.

Keywords

• Solanum lycopersicum
• Dwarfism
• Backcrossing
• Computational intelligence
• Genetic dissimilarity