Serum protein fingerprinting coupled with artificial neural network distinguishes glioma from healthy population and brain benign tumo

liu jian,wu minliang,dong qi,yu jiekai,hu yue,xu wenli,shen hong,zheng shu

DOI: https://doi.org/10.3760/j:issn:1009-9158.2005.01.009

2005-01-01

Abstract:objective To screen and evaluate protein biomarkers for the detection of glioma from healthy population and from brain benign tumor by using surface enhanced laser desorption/ionization mass spectrometry (SELDI MS) coupled with an artificial neural network algorithm (ANN) Methods SELDI TPF MS protein fingerprinting of serum from 105 patients of brain tumors and healthy men, including 28 patients with glioma, 37 patients with brain benign tumor, and 40 age matched healthy men were included in the study Two thirds of samples of total number of every compared pair were used to set up discriminating patterns One third of samples of total number of every compared pair were used to cross validate, simultaneously The discriminate cluster analysis derived SPSS10 0 software was used to compare gradeⅠ Ⅱgliomas with grade Ⅲ Ⅳones Results A blinded test showed a accuracy of 95 7%, a sensitivity of 88 9%, a specificity of 100%, a positive predictive value of 90% and negative predictive value of 100% when comparing the gliomas with healthy men A accuracy of 86 4%, a sensitivity of 88 9%, a specificity of 84 6%, were obtained when comparing the gliomas with benign brain tumors A total accuracy of 85 7%, a accuracy ofⅠ Ⅱgloma was 86 7%, a accuracy of Ⅲ Ⅳgloma was 84 6% when comparing the grade with grade Ⅲ Ⅳones (discriminate cluster analysis) Conclusion SELDI MS combined with bioinformatics tools can greatly facilitate the discovery of new and better biomarkers The high sensitivity and specificity achieved by the use of the selected biomarkers show great potential application for the discrimination of gliomas from healthy person and from brain benign tumors