Combining Gene Expression Datasets Using Generalized Singular Value Decomposition
Author Information
Author(s): Andreas W. Schreiber, Neil J. Shirley, Rachel A. Burton, Geoffrey B. Fincher
Primary Institution: Australian Centre for Plant Functional Genomics, University of Adelaide
Hypothesis
Can the generalized singular value decomposition effectively merge gene expression datasets with partial overlap?
Conclusion
The generalized singular value decomposition is a viable method for integrating gene expression datasets with only partial overlap, enabling the discovery of co-expressed genes.
Supporting Evidence
- The GSVD allows for the identification of genes present in only one dataset that are co-expressed with a target gene in the other dataset.
- The method is particularly useful for integrating large publicly available microarray datasets with smaller in-house expression measurements.
- The study demonstrated that co-expression analysis can lead to the identification of candidate genes involved in specific biological processes.
Takeaway
This study shows a way to combine two different methods of studying genes to find out which ones work together, even if they don't have all the same information.
Methodology
The study used generalized singular value decomposition to analyze and merge datasets from microarray and quantitative real-time PCR.
Potential Biases
Potential biases due to platform-specific systematic errors and differences in gene representation between datasets.
Limitations
The datasets had only partial overlap in gene content and experimental conditions, which complicated the analysis.
Participant Demographics
The study focused on barley (Hordeum vulgare L) tissues.
Statistical Information
P-Value
p<0.05
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website