figshare
Browse
Figure_1.tif (1.66 MB)

Multivariate discriminant analysis of the UBE2D2 normalised gene GeXP hCellMarkerPlex data from human colon normal (white triangle) (n = 24), adenomatous polyp (grey triangle) (n = 17) and carcinoma (black triangle) (n = 19) tissues.

Download (0 kB)
figure
posted on 2014-11-25, 03:09 authored by Janice E. Drew, Andrew J. Farquharson, Claus Dieter Mayer, Hollie F. Vase, Philip J. Coates, Robert J. Steele, Francis A. Carey

Information on the gene symbols on the biplot is available in Table S2. (A) Principal component analysis (PCA) biplot permits visualisation of inherent clustering patterns of individual tissue samples and associated gene expression levels. (B) Orthogonal Partial Least Squares Discriminant Analysis (OPLS-DA) was applied to fit a 2-class supervised model maximising covariance and discriminating gene expression profiles associated with the different tissues sample types; the biplot shows scores and loadings as well as the regression coefficients best explaining each class ($M4.DA(N),$M4.DA(A),$M4.DA(C). (C) Rank of importance of cell marker genes within the OPLS-DA. (D) Matrix showing the associated misclassification rates.

History