Impact Analysis: Beyond Traditional Enrichment

Impact Analysis is Advaita’s proprietary pathway analysis approach that goes beyond traditional enrichment analysis methods by identifying significantly impacted pathways based on two forms of evidence: Over Representation Analysis and Perturbation Analysis. Unlike all other pathway analysis methods that treat pathways as simple sets of genes, Impact Analysis accounts for the biological processes the pathways are designed to represent. Impact Analysis is the only pathway analysis method that uses a systems biology approach by leveraging information about the type, function, position, and interactions of genes within a given pathway.

Impact Analysis vs. Traditional Enrichment Analysis

Traditional pathway enrichment analysis methods have several key limitations. First, they calculate a p-value using a statistical model that assumes all genes are independent, a premise that is clearly flawed. The very purpose of a pathway is to represent the complex dependencies between genes involved. Ignoring these well-documented interactions and assuming genes to be independent undermines the validity of the analysis results.

Another key limitation is that traditional enrichment methods completely overlook the position and functional roles of genes within pathways, as well as the direction and type of signals that connect them (such as activation or repression). Each interaction and dependency has been established through extensive experimental research, yet this rich biological knowledge is disregarded by simple enrichment approaches.

In a meta-analysis, iPathwayGuide™ has the highest median value of AUC, which, according to the authors of the report, “is the most comprehensive and important (metric) because it combines both the sensitivity and specificity across all possible thresholds.” iPathwayGuide™ demonstrated it was more successful identifying the knockout gene compared to other tools.

In contrast, Impact Analysis constructs a mathematical model represented by a system of equations that captures every interaction within a pathway. This model accounts for the position and role of each gene, the direction and type of signals between genes, and any feedback loops present. Using this system of equations, Impact Analysis calculates perturbations at the gene level, which are then integrated to quantify the overall pathway perturbation. This approach enables accurate identification of truly impacted pathways while reducing false positives, providing more biologically meaningful and reliable analysis results.

“We handle a lot of big data. The results from our experiment aren’t just a single gene or a single pathway change, but more widespread changes that happen across the entire genome. When you have hundreds or thousands of different, in my case, RNASeq data, iPathwayGuide™ allows us to make sense of the data without having to go through each individual gene and determine what its function is, and how the genes are related.”

– Britney Helling, Ph.D, Carole Ober’s laboratory at The University of Chicago

“…you can add chemicals that interact with it. You can add upstream regulators, all these things that you can very quickly get to a plot that allows you to see your results in a network form that is insightful and meaningful, and not just a huge pathway diagram that just overwhelms people.”

– Michael Chimenti, Ph.D, University of Iowa Bioinformatics Division