Sentence examples for discriminating line from inspiring English sources
MFSelector does not count samples on the wrong side of the discriminating line more than once.
The number of samples on the wrong side of the discriminating line is referred to as "discriminating errors".
Thus, we can get the "discriminating errors" for "Level One" process by drawing a discriminating line through one of the samples in Stage One to determine the number of samples in Stage One that have lower expression values than the discriminating line and the number of samples in other stages having higher expression values than the discriminating line.
Note that, when more than one discriminating line for Stage One have the same number of discriminating errors, we can choose any one of the lines; here we use the discriminating line corresponding to the lowest sample number.
For example, one sample from Stage One and another sample from Stage Four are on the wrong side of the discriminating line when considering the 2nd sample in Stage One to draw the discriminating line, resulting in two discriminating errors.
The discriminating line for each level is also selected in the same manner as that for the ascending case.
The linearity of AQAMA assay for LINE-1 methylation level (Figure 1) was evaluated by Pearson's coefficient of linearity: 0.99 (P<0.001), which showed high accuracy in discriminating LINE-1 methylation level differences.
Figure 6(B) displays DNMT3B, one of the top 12 descending genes with DEtotal = 0. Here, samples from all stages are also well separated by discriminating lines for each level.
LIN28A is without N-1 distinct discriminating lines because its expression is tightly grouping from Stage One to Stage Three.
Using DEtotal <= 7 with the constraint p-value<1.0E-5 and with N-1 distinct discriminating lines (here N = 5), we get 857 ascending and 1,117 descending monotonic genes.
In this case, we obtain 7, 2, and 2 discriminating errors taking each of three samples in order from Stage One to draw the discriminating lines.
