Figure 1.

Workflow of this study. The DNA microarray gene expression datasets for the four serous carcinoma (SC) staging groups and normal ovarian controls were downloaded from the publicly available database. The gene set regularity (GSR) index was computed by measuring the changes of the gene expression ordering of the gene elements in the Gene Ontology (GO) gene set. The functionome consisting of 5917 GO gene set defined functions was reconstructed for each sample. Then, the immunofunctionome consisting of 333 immunological functions was reconstructed by extracting the immune-ancestor GO terms from the functionome for the four staging and normal control groups. Machine learning was applied to recognize the patterns of the functionomes and then executed the binary and multiclass classifications. The key immunological functions were extracted by the statistical methodology and a series of filters from the immunofunctionomes.