OmicsAnalyst

Meta-analysis: upload studies

Combine several studies that measured the same omics type on different samples. Each study keeps its own samples and metadata; the analysis runs within each study first and is then combined across studies, so differences in platform or scale between studies are handled.

1. Shared settings

These apply to every study: a meta-analysis combines measurements of the same features. If a study uses different IDs, change its ID type in the study table; all studies are mapped to Entrez IDs.

Omics type

Organism

Feature ID type (default for all studies)

2. Studies

One table per study (features × samples), or one combined table split by a study column.

Study	File	Samples	Features	ID type	IDs mapped	Value scale

No studies yet

Add at least two studies of the same omics type, or try an example — genes (three mouse helminth-infection studies) or metabolites (four GC-TOF-MS adenocarcinoma studies).

Helminth infection [3]	Three mouse transcriptomics studies (microarray) of helminth infection: control vs infected (38 samples in total, about 5,000 genes each). From GEO: GSE25713, GSE59276, GSE69588.	Study 1 Study 2 Study 3
Lung adenocarcinoma [4]	Four untargeted GC-TOF-MS blood metabolomics data sets: control vs lung adenocarcinoma (150–180 metabolites each, matched across studies by name), from two patient cohorts (Miyamoto, S. et al.). Data sets 1 and 3, and 2 and 4, contain the same samples. This is the statistical meta-analysis example from MetaboAnalyst.	Study 1 Study 2 Study 3 Study 4
Tuberculosis [3]	Three human whole-blood transcriptomics studies: latent vs active tuberculosis (152 samples, 17,837 genes), from the UK, South Africa and The Gambia, on two microarray platforms, with each sample's sex. The study information (country, platform) is filled in for "Why do the studies disagree?" on the results page. From GEO (GSE19439/GSE19442, GSE28623), as used by Sweeney, T.E. et al.	Study information One study (.gz)