Building a classifier for kidney tumor diagnosis

Knowing that the prediction ability of the various classifiers was high, we applied the same methods to all samples in order to build a final classifier that could be used for diagnostic purposes. In the following, we will concentrate on the classifier of the PAM method.

The classifier of the PAM method was based on 18 genes. We visualized the generalized log ratios of these genes by different colors (Plate 2). Among the 18 genes were GAPD, CXCL14, ADFP, SPP1, and CD9. Several of these are known key players in cancer and differentiation: CXCL14 is a cytokine that is frequently down-regulated in tumors (24) but up-regulated in inflammatory cells of the tumor microenvironment (25), ADFP is a protein involved in cell differentiation and has been found to be highly over-expressed in ccRCC (26). The osteopontin gene (SPP1) is a target for TP53 (27) and a lead marker for colon cancer progression (28). Two members of the galactose-binding lectin family (LGALS3 and LGALS9) are known to be involved in tumorigenic processes. Two genes (CD9 and TSPAN1) coded for cell surface proteins of the tetraspanin family, which mediate signal trans-duction events in cell development, activation, growth and motility. This indicates that the three kidney tumor types can be distinguished by the activation status of specific biological processes. Two of the classifier genes (GAPD and LGALS3) were represented by independent clones. The gene expression patterns of these clones were very similar, suggesting that different spots on the microarrays give highly reproducible results.

The relative gene expression pattern suggested that the chRCC tumors can be clearly distinguished from the ccRCC and pRCC. With the exception of GAPD, there is an almost inverse relationship of expression patterns of all 18 genes between these two groups. In contrast, the distinction between ccRCC and pRCC tumors is not obvious and appears to rely primarily on the expression of the GAPD gene, which is up-regulated in ccRCC but down-regulated in pRCC. This may reflect different metabolic activities between these two tumor types. This finding is consistent with a previous report about high levels of glycolytic enzymes in ccRCC samples (29). The importance of GAPD as a part of the classifier was supported by its low expression in tumor 14, which was consistently misclassified in all algorithms. Apart from GADP, ccRCC and pRCC tumor types differ mainly by the expression of crystallin alpha B (CRYAB), osteopontin (SPP1) and tetraspanin 1 (TSPAN1).

10 Ways To Fight Off Cancer

10 Ways To Fight Off Cancer

Learning About 10 Ways Fight Off Cancer Can Have Amazing Benefits For Your Life The Best Tips On How To Keep This Killer At Bay Discovering that you or a loved one has cancer can be utterly terrifying. All the same, once you comprehend the causes of cancer and learn how to reverse those causes, you or your loved one may have more than a fighting chance of beating out cancer.

Get My Free Ebook

Post a comment