At present the machines are ruling the world by dumping it with awash data. Very often, the sensors like instruments are used to monitor various live factors. It results in exponential growth of time series data over past few years. Excavating the patterns from these time series data can be used for future value prediction, behavioural study and so on. Clustering is one of the unsupervised pattern mining techniques. Due to increase in dimensionality of data, the biclustering approach will better suits to time series data than that of the clustering approach. Further optimal patterns can be mined using the optimized biclustering approach i.e. Genetic Algorithm based biclustering. To overcome the big data complexity MapReduce model is chosen besides the Genetic Algorithm. For Optimal pattern mining from the time series data a novel approach MapReduce based Genetic Algorithm for Biclustering is proposed. This MR-GABiT is attempted to the Yeast cell cycle microarray time series gene expression data. This proposed approach mines highly correlated patterns from the gene expression data.