Skip to main content
Fig. 1 | BioData Mining

Fig. 1

From: Principal component analysis based unsupervised feature extraction applied to budding yeast temporally periodic gene expression

Fig. 1

Upper triangle: scatter plots in YMC between the kth (1≤k≤4) PC loadings, v kj −〈v kj 〉 j ,1≤j≤M, where M is the number of time points. Lower triangle: winding number W(M−1); W(M−1) at the kth row and k ′(<k)th column correspond to the scatter plot at the symmetric position, i.e., at the k ′th row and kth column. Time dependence of PC2 and PC3 loadings are shown in Fig. 3 b. The contributions of PC loadings are 88.7, 6.2, 2.1 and 0.6 %, respectively, indicating that PC loadings,v kj ,k=2,3, with contributions as small as a few percent that correspond to tens of genes when the total number of genes correspond to thousands, cannot be disregarded

Back to article page