權重型基因法則於肝癌質譜資料特徵選取之應用

Application of Gene Weighted Genetic Alogorithm to feature selecte with SELDI-TOF MS Hepatoma data

潘咨亦(Tzu-Yi Pan)

指導教授 ： 姚立德

國立臺北科技大學/機電學院/機電整合研究所/碩士(2007年)

https://doi.org/10.6841/NTUT.2007.00408

摘要

對於一份經由表面增強的雷射解析電離飛行時間質譜儀（Surface Enhance Laser Desorption/Ionization Time-Of-Flight Mass Spectrometry, SELDI-TOF MS）並且搭配該儀器生產公司所附的Ciphergen Protein Chip 軟體產生的我國國人臨床肝癌峰值資料，想要將其做患有癌症與否的分析是有著相當程度的困難。雖然已經藉由軟體大幅降低所需分析的資料量和雜訊，但是卻衍生出其他的問題。所以本文先提出由生物醫學之觀點，將資料轉換為有相同依據的特徵向量。再以此作為分析上的依據，使用特徵選取搭配支援向量機做資料的分類。
將基因法則(Genetic Algorithm)應用於特徵選取上則是近年來被發展出來的方法。而本文所提出的權重型基因法則(Gene Weighted Genetic Algorithm， GWGA)，將染色體中的每一個基因給予一個交配的權重値，並以機率的方式來做交配的運算，改變了傳統基因法則交配的方式以解決容易陷入局部最佳解(local optimum)的問題，並且能夠適當的減少特徵數目。

關鍵字

SELDI-TOF MS ； 支援向量機 ； 特徵選取 ； 基因法則

並列摘要

It’s very difficult to analyse the clinical SELDI-TOF MS （Surface Enhance Laser Desorption/Ionization Time-Of-Flight Mass Spectrometry）data which is obtained form our countrymen. The data also had been pre-processed with the software which is appended with the instrument producer. Although the data have already been reduced the capacity of data and noise, come into another problems. In this way, it is proposed to transform the data into feature vectors on equal basis which according to the biomedical point in this dissertation. Using feature selection with SVM（Support Vector Machine） to classifer these feature vectors datas.
Among the different categories of feature selection algorithms, the genetic algorithm (GA) is a rather recent development. In this dissertation, the Gene Weighted Genetic Algorithm, that is give a gene weighted value in each gene of chromosome, and make the crossover operation by the probability, this change can reduce local optimum when used traditional method of Genetic Algorithm.

並列關鍵字

SELDI-TOF MS ； Support Vector Machine ； Feature Selection ； Genetic Algorithm

參考文獻

[41] 西滿正，癌的最新診斷與治療，台北：建宏，1996。

[6] G. Bill, S. Mian, F. Holding, R. O. Allibone, J. Lowe, S. Ali, G. Li, S. McCardle, I. O. Ellis, C. Creaser, and R. C. Rees, “An integrated approach utilizing artificial neural networks and SELDI mass spectrometry for the classification of human tumours and rapid identification of potential biomarkers,” Bioinformatics, vol. 18, 2002, pp. 395-404.

[8] R. H. Lilien, H. Farid, and B. R. Donald, “Probabilistic disease classification of expression -depentent proteomic data from mass spectrometry of human serum,” J. of Computational Biology, vol. 6, 2003, pp. 925-946.

[9] E. F. Petricoin, A. M. Ardekani, B. A. Hitt, P. J. Levine, V. A. Fusaro, S. M. Steinberg, G. B. Mills, C. Simone, D. A. Fishman, E. C. Kohn, and L. A. Liotta, “Use of proteomic patterns in serum to identify ovarian cancer,” Lancet, vol. 259, 2002, pp. 572-577.

[10] L. H. Loo, J. Quinn, H. Cordingley, S. Roberts, L. Hrebien, and M. Kam, “Classification of SELDI-TOF mass spectra of ovarian cancer serum samples using a proteomic pattern recognizer,” IEEE 29th Annual, proceed. of Bio. Eng., 2003, pp. 130-131.