Example for multi class classification using One vs All method with SVM

I need to classify data into 4 classes. I intend to use the One vs All method with SVM to do so. I have a TrainingData matrix (Dimension: (400x4), 400 records, each having 4 features) and a Label matrix (Dimension: (400x1), having values {1,2,3,4}).

I would be highly appreciate it, if someone could post the rough schematics of MATLAB code (how to call the 'svmtrain' functions) in order to achieve this.

Thanks in advance.