Meta-analysis is a type of statistical technique used to integrate multiple independent datasets that have been collected to study same or similar experimental conditions, in order to obtain more robust biomarkers. By combining multiple data sets, the approach can increase statistical power (more samples) and reduce potential bias.
A key concept in meta-analysis is that it is generally unadvisable to directly combine different independent datasets (i.e. merge them into a single large table) and analyze them as a single unit. This is due to potential batch effects associated with each datasets, which can completely overwhelm the biological effects. This issue has been well-studied in microarray experiments generated from different platforms. It is expected the issue could be more severe due to the lack of standardization in metabolomics.
Instead, meta-analysis is usually computed based on summary statistics (p values, effect sizes, etc.) to identify robust biomarkers. The Statistical Meta-analysis module in MetaboAnalyst was developed to support these approaches. The results can be visualized using heatmap to explore the patters across different studies.