Data-Mining Techniques: A New Approach to Identifying the Links among Hybrid Strains of Pleurotus with Culture Media

Abstract

In this study, a data set of mycelial and cultural characteristics of hybrid strains of Pleurotus ostreatus and Pleurotus djamor were analyzed using three data-mining techniques: the K-medoids clustering algorithm, PCA biplot and the association rules algorithm. The characteristics evaluated were as follows: maximum velocity; lag phase; biomass; and exopolysaccharides content in the cultivation of 50 hybrid strains of Pleurotus ostreatus and 50 hybrid strains of Pleurotus djamor. Different mixtures of culture media were used to supplement Ecuadorian agricultural products. Data of the parameters obtained in the experimental methods were grouped into four clusters, obtaining a presentation of the hybrid strains of Pleurotus with a higher relation to each characteristic measured. Data-mining tools showed the hybrid strains cultivated on solid-culture media (M1 = malt extract agar and rice flour) and liquid-culture media (L1 = maltose, yeast extract and rice flour) presented the highest mycelial and cultural characteristics. These results are good indicators to improve the industrial production of edible fungi by using rice flour in the cultivation, contributing to the mushroom market and circular economy.