Skip to main content
Figure 6 | BMC Bioinformatics

Figure 6

From: Principal component analysis for predicting transcription-factor binding motifs from array-derived data

Figure 6

Leave-one-out test. (A) Model error with a randomized promoter matrix. The mean and s.d. of sum square error is 108 ± 31 (N = 1,000), and the label "a" indicates the error with the original promoter sequences. (B) Model error with randomized gene expression ratios. The mean and s.d. of sum square error is 93 ± 23 (N = 1,000), and the label "b" indicates the error with the original gene expression ratios.

Back to article page