With the rapid increase in the size of microarray datasets, identifying functionally related genes from large and diverse microarray gene expression datasets is a key challenge.
Although a perfect or the best solution to it is yet to be found, we have demonstrated that the ISML pipeline can reduce the dimensionality of microarray datasets, identify and rank classifier genes, generate a small set of classifier genes, produce an SVM classification model with high accuracy, and select a small group of biomarker candidate genes for biological validation.
Comparison of the A. suum TES transcriptome with the C. elegans microarray dataset identified 165 A. suum germline-enriched genes (83% are spermatogenesis-enriched).
Using our proposed meta-analysis method on the two microarray datasets, we identified 330 probe sets representing 309 genes that were significantly relevant to CAN.
Through comparative analysis of a number of microarray datasets we identified a set of genes commonly regulated by FOXO proteins and PI3K PKB/Akt, which includes CTDSP2 (C-terminal domain small phosphatase 2).
Recently Reis et al. [ 13] have performed a meta-analysis based on five publicly available microarray datasets, and identified a four-gene signature that is of prognostic value for oral cancer reoccurrence.
The results of our analysis are summarized in the following major steps: (1) data acquisition and merging of multiple microarray datasets to identify DEGs in tumor samples, (2) analysis of the gene coexpression network to identify cancer stage-associated modules, and (3) identification of cancer hub genes and development of a key hub gene-based classifier model to distinguish OSCC stages.
We analyzed several independent microarray datasets, to identify a predictor, use it to classify unclassifiable sarcomas, and assess oncogenic pathway activation and chemotherapy response.
In order to find control genes for comparing human-chimpanzee gene expression across multiple tissues, we developed a pipeline that draws on information from published microarray datasets for identifying candidate normalizers.
Comparative analysis of microarray datasets can identify patterns and consistencies not discernable from any one dataset individually.
