The table shows the fraction of feature-associated patterns for each peptide set and the respective number of patterns associated with each feature.
The fraction of features removed at each iteration, 10% in our experiments, greatly influences the computational complexity of the method.
This procedure was iterated, and in each iteration we computed the fraction of features with positive correlation coefficients.
Assuming that the underlying signal is sparse, i.e. that only a fraction of features contribute to a principal component (PC), this estimation consistency can be retained.
Even without knowing the exact statistic for the DFC-score, it can be used for ranking features and selection of a fixed number, or best fraction of features with highest score.
When a high fraction of features, ν > 0.5, is taken as differentially expressed the difference between the DFC test and t-test is minimal, as both tests remove the most easily detectable, non-expressed features.
To select significant features from FDAFSA, we trained our model using a different fraction of features than the hexamer_set k of training_data k and tested our model with test_data k. Figure 1 shows the average sensitivities and specificities of different fractions of 4,096 features.
Normalization methods for mRNAs often rely on one or more of the following assumptions: only a small fraction of features are differentially expressed and the proportions of up- and down-regulated expressions are approximately equal.
When a very small fraction of features ν ~ 1/ Np is taken as differentially expressed, resulting in only few features selected, the difference between the DFC test and t-test will be small again, as the differential expression of the few features should be very strong and can be effectively selected by t-test alone.
As discussed previously, unlike mRNA datasets, where it has been well established that only a small fraction of features can be differentially expressed; for miRNA datasets, we could not favor, a priori, any region of (0,1) for ξ →, because the proportions of differentially expressed features among different miRNA datasets could vary dramatically.
