Biomarker Identification for Prostate Cancer and Lymph Node Metastasis from Microarray Data and Protein Interaction Network Using Gene Prioritization Method
Table 6
Data and Text Mining Gene Prioritization Methods.
Method
Brief description
Reported results
Gene seeker
Gathers gene expression and phenotypic data from human and mouse from nine databases. Relies on the assumption that disease genes are likely to be expressed in tissues affected by that disease [6]
Offers a web-service to find disease-related genes to the input genetic localisation and phenotypic/expression terms
eVOC
Co-occurrence of disease name on PubMed Abstracts. It selects the disease genes according to expression profiles [5]
It was tested on 417 candidate genes, using 17 known disease genes. It successfully retrieved 15 of the 17 known disease genes and shrunk the candidate set by 63.3%
It accomplished a performance of up to 0.90 in their ROC curve.
Limitations
Generally imposed by the source data which carries little knowledge about the disease. For instance GO terms include brief description of the corresponding biological function of the genes but only 60% of all human genes have associated GO terms, and they may be inconsistent due to differences in curators’ judgement [16]