Journal: Neural Plasticity

Article Title: Classification of Visual Cortex Plasticity Phenotypes following Treatment for Amblyopia

doi: 10.1155/2019/2564018

Figure Lengend Snippet: Identifying plasticity features using the principal component analysis. (a) .The percentage of variance captured by each principal component by singular value decomposition (SVD) applied using all of the protein expression data. The first 3 principal components capture 54%, 18%, and 10% of the variance, respectively, totalling >80% and thus representing the significant dimensions. (b). The quality of the representation, cos 2 , for the proteins is plotted for each dimension (small/white: low cos 2 ; large/blue: high cos 2 ). (c). The sum of cos 2 values for the first 3 dimensions for each protein. (d, e). Biplots of PCA dimensions 1 and & 2 and (f, g). 1 and & 3. These plots show the vector for each protein (d, f) and the data (small dots) plus the average (large dots) for each condition with the best-fitting ellipse (e, g). (h). The basis vectors for dimensions 1-3 showing the amplitude of each protein in the vector. (i). The strength (circle size) and direction (blue-positive, red-negative) of the correlation ( R 2 ) between each protein and the PCA dimensions. (j). Correlation between the plasticity features (columns) identified using the basis vectors (see Results) and then PCA dimensions 1-3. Filled cells are significant, Bonferroni- corrected correlations (green = positive, red = negative). For the table of Pearson's R values and significant p - values for these associations, see Supplemental .

Article Snippet: The data were centered by subtracting the mean column vector and applying singular value decomposition (SVD) to calculate the principal components (RStudio).

Techniques: Expressing, Plasmid Preparation

Journal: Heredity

Article Title: Genomic selection using principal component regression

doi: 10.1038/s41437-018-0078-x

Figure Lengend Snippet: Comparison of computing times for three methods of eigenvalue decomposition for an X matrix with 5000 rows and 100,000 columns

Article Snippet: However, the common practice in PCR is to use singular value decomposition (SVD) for the feature matrix X n×m as (Bordier et al. 2011 ; Mandel 1982 ; Shlens 2014 ) X n × m = U n × n Δ n × m V m × m T , 2 where the columns of U are called the left singular vectors, the columns of V are called the right singular vectors and Δ is a rectangular diagonal matrix with non-negative real numbers on the diagonal.

Techniques: Comparison