S. Mahalakshmi, R. Pangayar Selvi, V. Anandhi and N. Bharathi
Abstract
The tolerance of plants to environmental stress is aided by numerous proteins, one such protein is a LateEmbryogenesis Abundant (LEA) Protein. Thus, it is very essential to identify and classify LEA proteins.For this purpose, machine learning would be greatly useful. In this study, an attempt was made to developSVM classifiers for the classification of LEA proteins based on their physico-chemical properties. The physico -chemicalproperties were analyzed from protein sequences and then the dataset is split into training andtest data. The model is developed and evaluated using cross-validation techniques like 5-fold. The resultsshow that the accuracy of developed SVM classifiers ranges from 60 to 97 per cent.