An ABC algorithm is developed as described in Section 2.1 to select the most informative and predictive genes from an mRMR dataset that give the highest classification accuracy with an SVM classifier.
Previous computational methods identified different sets of histone marks to be the most predictive of mRNA production; Karlic et al. found H3K4me3 and H3K79me1 to be most informative in predicting the expression level in low CpG content promoters whereas the expression in high CpG ones depend more on H3K27ac and H4K20me1 [ 20].
Visualizing the class distribution seems to be the most informative for determining the predictive accuracy in these scenarios.
The finding that the GHmax value from the spontaneous profile was the most informative variable, and more predictive than the GH peak in the provocation test, was known previously [ 1], and reflects both the higher reproducibility of the former and its greater relevance to growth [ 31, 32].
The context++ model was more predictive than any published model and at least as predictive as the most informative in vivo crosslinking approaches.
However, choosing the optimal and most informative phenotyping approach to discover novel and predictive biomarkers for patient cohorts has become a major challenge and greatly hampers knowledge gain to successfully develop and tailor new and existing therapies to suitable patient collectives [1].
Such feature identification can also be used to help streamline future prediction efforts [by focusing efforts on those predictive features that are most informative for the task at hand (Ko and Lee 2009)].
For this example, we consider the predictive approach to produce the most informative risk picture, as it is the patient perspective that is of concern.
As expected, the highest predictive accuracy was obtained under the most informative setting S1, whereas the practically more realistic settings resulted in reduced accuracies (Tables 2 and 3).
Indeed, we demonstrated that different fusion types exhibited different predictive values on biopsy outcomes; the most informative fusion markers composed of TMPRSS2 ERG subtype III, IV or any TMPRSS2 ETS (ETV4, ETV5, ETV5) fusions, a special combination we termed Fx (III, IV, ETS).
