The Integration of Quantitative and Multicategorical Data for the Analysis of Genetic Divergence in Germplasm of Cassava (manihot Esculenta Crantz.): A New Approach

The analysis of genetic divergence in plant germplasm based on a set of quantitative data has revealed a low association to the divergence estimated from multicategorical data and vice versa.On the other hand, strategies involving the integration of quantitative and multicategorical information may provide more accurate estimations.Thus, this study aimed to estimate the genetic divergence in the germplasm of cassava based on the integration of quantitative and multicategorical data, targeting a greater comprehensiveness and accuracy in the estimation.Data from 10 genotypes of M. esculenta, characterized by 20 quantitative and 24 multicategorical characteristics was used in this work.The genotypes were collected from autochthonous fields from five different microregions of Brazil and evaluated in a completely randomized block design experiment with four replications.First, the analysis of genetic divergence was performed based on the distances individually obtained from the quantitative and multicategorical information.For the data integration, three strategies were adopted, the Gower's algorithm, the transformation of quantitative data by the equitable division of its amplitude into several classes, and the sum of the distance matrices, obtained from quantitative and multicategorical data.The estimates of genetic divergence from the quantitative data had a low association with that expected from the multicategorical data and vice versa.The transformation of data and the algorithm of Gower were not efficient, which resulted in low correlations with the matrices of distances obtained from the original data.The sum of matrices consisted on the strategy of higher efficiency and provided a higher comprehensiveness and accuracy in the analysis o f genetic divergence in the germplasm of cassava.