This combined approach enhanced the quantitative and kinetic understanding of apoptosis signal transduction, but also provided new insights that systems-emanating functions (i.e., functions that cannot be attributed to individual network components but that are instead established by multi-component interplay) are crucial determinants of cell fate decisions.
We call this problem the enumeration of system phenotype-determining component interplays.
Thus far, we have presented how to detect component interplays from an instance-based data.
Finally, the ensemble of classifiers is formed to predict the system's phenotype(s) given the values of all its component-interplay groups (Step 5: bringing component interplays altogether section).
We propose an algorithm, named Spice, to address the new problem of enumeration of system phenotype-determining component interplays.
The candidate component-interplay group identified in Steps 1-3 is probably not the only group of system components that is responsible for a system's behavioral phenotypic state.
Here, we test three different weighting schemes described in Step 5: bringing component interplays altogether section: majority voting, training accuracy-based voting, and internal cross-validation-based voting.
In this paper, we addressed the important and challenging problem of enumerating statistically significant and application-relevant component interplays that are key contributors to the system's phenotype.
By bringing the enumerated component interplays altogether (Step 5) a good ensemble of classifiers can be achieved (as illustrated in Results and discussion section).
At a higher level, Spice first identifies a candidate component (feature) set (Step 1: identifying candidate component interplays section), it then scores its phenotype specificity-determining skill (Step 2: scoring candidate component interplays section) along with statistical significance assessment (Step 3: assessing statistical significance section).
These three steps are repeated in an iterative fashion by "knocking out" the selected candidate component sets until the stopping criterion is met (Step 4: iterative "knock-out" of component interplays section).
