A combined selection approach using modified multivariate analysis for identification of fast cooking beans (Phaseolus vulgaris L.) based on seed physical parameters

The present study reports comparative advantage of a modified multivariate analysis based on seed physical parameters related to cooking in beans. The modified GC*T biplot represents functional relationship of cooking time score with seed physical traits. Compared to normal GT based principal component analysis (PCA) that concentrated variability in first six principal components (83.40 per cent) the GC*T concentrated the variability in first three principal components (91.63 per cent) with first two accounting for 88.40 per cent variation. In terms of cooking time score, WB-216 and WB-1518 were easy to cook (5) followed by WB-1678, GLP-1, WB-1634, WB-N-14, WB-869, WB-957, WB-501, WB-22, WB-2020-180 and WB-6 (CTS 4). Eighteen genotypes had longer cooking times and eight were intermediate in cooking time. Based on mean superiority index (MSI) WB-216 had highest MSI value (2.14) followed by WB-1518 (1.45) while as highest negative MSI values were recorded for WB-371 (−1.005). The GC*T approach used helps us translate seed physical traits into faster cooking trait making the GC*T approach more meaningful as compared to GT based on traits per se.