ABSTRACT
Statistical meta-analysis is a process of integrating multiple related studies with an objective to generalize the results of single study experiments with more statistical power. Meta-analysis in big data fields such as bioinformatics have been recognized as powerful methods to obtain robust estimates of interest. In this chapter, an effort has been made to present methods used for transcriptomic data like DNA microarray, and RNA-seq.
Gene expression technologies suffer from a major design issue – expression of hundreds of thousands of genes has to measure in a small number of samples. In addition, biological and technical variability coupled with high experimental cost of these studies necessitate the need to develop methods of statistical meta-analysis in bioinformatics.
The procedure of meta-analysis is based on the end objectives of study; however, I attempted to present some generic steps which are typical to any meta-analysis.
In short, meta-analysis is a very powerful technique to derive new hypotheses and knowledge from high-throughput biological data. However, the validity of meta-genes should be checked using low-throughput experiments such as real time PCR, western blot, immunohistochemistry etc.
