Table of Contents Author Guidelines Submit a Manuscript
Disease Markers
Volume 29, Issue 5, Pages 231-242
http://dx.doi.org/10.3233/DMA-2010-0753

The PCA and LDA Analysis on the Differential Expression of Proteins in Breast Cancer

Seng Liang,1 Manjit Singh,2 Saravanan Dharmaraj,1 and Lay-Harn Gam1

1School of Pharmaceutical Sciences, Universiti Sains Malaysia, Penang, Malaysia
2Department of Surgery, Penang General Hospital, Penang, Malaysia

Received 31 December 2010; Accepted 31 December 2010

Copyright © 2010 Hindawi Publishing Corporation. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract

Breast cancer is a leading cause of mortality in women. In Malaysia, it is the most common cancer to affect women. The most common form of breast cancer is infiltrating ductal carcinoma (IDC). A proteomic approach was undertaken to identify protein profile changes between cancerous and normal breast tissues from 18 patients. Two protein extracts; aqueous soluble and membrane associated protein extracts were studied. Thirty four differentially expressed proteins were identified. The intensities of the proteins were used as variables in PCA and reduced data of six principal components (PC) were subjected to LDA in order to evaluate the potential of these proteins as collective biomarkers for breast cancer. The protein intensities of SEC13-like 1 (isoform b) and calreticulin contributed the most to the first PC while the protein intensities of fibrinogen beta chain precursor and ATP synthase D chain contributed the most to the second PC. Transthyretin precursor and apolipoprotein A-1 precursor contributed the most to the third PC. The results of LDA indicated good classification of samples into normal and cancerous types when the first 6 PCs were used as the variables. The percentage of correct classification was 91.7% for the originally grouped tissue samples and 88.9% for cross-validated samples.