| Literature DB >> 23722107 |
Lei Cheng1, Qing Zhang, Sheng Yang, Yanqing Yang, Wen Zhang, Hengjun Gao, Xiaxing Deng, Qinghua Zhang.
Abstract
A widely held viewpoint is that the use of multiple markers, combined in some type of algorithm, will be necessary to provide high enough discrimination between diseased cases and non-diseased. We applied stepwise logistic regression analysis to identify the best combination of the 32 biomarkers at chromosome 8q on an independent public microarray test set of 80 paired gastric samples. A combination of SULF1, INTS8, ATP6V1C1, and GPR172A was identified with a prediction accuracy of 98.0% for discriminating carcinomas from adjacent noncancerous tissues in our previous 25 paired samples. Interestingly, the overexpression of SULF1 was associated with tumor invasion and metastasis. Function prediction analysis revealed that the 4-marker panel was mainly associated with acidification of intracellular compartments. Taken together, we found a 4-gene panel that accurately discriminated gastric carcinomas from adjacent noncancerous tissues and these results had potential clinical significance in the early diagnosis and targeted treatment of gastric cancer.Entities:
Keywords: Biomarkers; Gastric cancer; Neoplasm invasion; Regression analysis; SULF1 protein, human
Mesh:
Substances:
Year: 2013 PMID: 23722107 DOI: 10.1016/j.ygeno.2013.05.004
Source DB: PubMed Journal: Genomics ISSN: 0888-7543 Impact factor: 5.736