Data mining for gene networks relevant to poor prognosis in lung cancer via backward-chaining rule induction

Abstract

We use Backward Chaining Rule Induction (BCRI), a novel data mining method for hypothesizing causative mechanisms, to mine lung cancer gene expression array data for mechanisms that could impact survival. Initially, a supervised learning system is used to generate a prediction model in the form of "IF <conditions> THEN <outcome>" style rules. Next, each antecedent (i.e. an IF condition) of a previously discovered rule becomes the outcome class for subsequent application of supervised rule induction. This step is repeated until a termination condition is satisfied. "Chains" of rules are created by working backward from an initial condition (e.g. survival status). Through this iterative process of "backward chaining," BCRI searches for rules that describe plausible gene interactions for subsequent validation. Thus, BCRI is a semi-supervised approach that constrains the search through the vast space of plausible causal mechanisms by using a top-level outcome to kick-start the process. We demonstrate the general BCRI task sequence, how to implement it, the validation process, and how BCRI-rules discovered from lung cancer microarray data can be combined with prior knowledge to generate hypotheses about functional genomics.

Keywords: C4.5; class discovery; data analysis; decision trees; microarray; molecular mechanisms; non-small cell lung cancer; semi-supervised methods; systems biology.

Figure 6.

A decision tree.

Figure 7.

Joint distribution of attribute A and Class

Figure 8.

Possible threshold values for a continuous attribute, C

Figure 1.

candidate split points.

Figure 2.

C45W-BCRI rules expressed as an AND/OR graph. *The rule to predict MRPL19 < 161.4 in the Low Risk trace will also predict MRPL19 < 161.4 in the High Risk trace. It is not shown again in the High Risk trace.

Figure 3.

Pathway Assist™ diagram showing SERPINA1 and ELA2 relationships of Example 1. The protein products are indicated by the large ovals, a binding interaction is indicated by the purple dot relationship between the ovals, and gene expression regulation is indicated by a square along a dotted line.

Figure 4.

Pathway Assist™ diagram illustrating FXN and EIF2S1 relationships of Example 2

Figure 5.

Pathway Assist™ diagram of KRT13 and DDX5 relationships of Example 3