| Objective By surface enhanced laser desorption/ionization time-of-flight mass spectrometry,serum protein fingerprinting from gastric cancer(GC) and healthy people be coupled with a pattern-matching algorithm were compared, then explore the potential clinical value. Methods Serum protein of 29 cases of patients with gastric cancer,34 cases after operation and 26 healthy people were tested by IMAC3 chip and proteinchip reader of Ciphergen Inc. Utilizing Ciphergen Proteinchip software analysis data in different groups,proceeding analysis of variance by Biomarker Wizard software of data of the same M/Z values protein in different group. Dividing data into different group and proceeding analysis of relativity by Biomarker Pattern software .Result At the key M/Z values range from1949.209 to 5070.169 Da,there were 18 protein contents in gastric cancer and healthy people groups are obviously different by the software analysis. And the classification rule of tree was discovered,in study model,analysis of the spectra from gastric cancer and healthy people corresponding correct ratio was 100%.In test model, correct ratio was 94.55%(52/55 ), sensitivity and specificity was 96.55%(28/29 )and 92.30%(24/26); At the M/Z values range from1871.658 to7566.822 Da, there were 32 protein contents of the 3 classes serum protein in gastric cancer , healthy people,post operation people groups are obviously different. In study modle, analysis of the spectra corresponding correct ratio was 95.50%(85/89),sensitivity and specificity was95.23% (60/63) and 96.15% (25/26) respectively. In test model, correct ratio was 84. 27% (75/89 ), sensitivity and specificity was 80. 95%(51/63 ) and 92. 30%(24/26 ). While the sensitivity and specificity ofcombined CEAand CA19-9 was23. 81% (15/63), and92.31% (24/26) respectively. Conclusion Compared with traditional tumor biomarker, gastric cancer can be quickly and exactly identified by SELDI-TOF mass without complex operation. So it shows amplitude clinical application.
|
PDF Full Text Request