Tech Notes

Q. How is the "Similar Lists" p-value calculated and how can I interpret it?

A. GeneSpring compares a set of genes with other existing gene lists in view of identifying significant representation of GO terms or other biologically relevant lists in respect to the entire genome.

For each gene list found that is similar to a selected gene list, a similar lists p-value or probability is reported. This statistic can be viewed under the "Similar Lists" tab in the Gene List Inspector.

The probability distribution on the bottom right displays the probability of expected overlap for two sets of 50 and 25 elements in size, each sampled randomly from a set of 100 elements.

Equation:

The probability of overlap corresponding to k or more genes between a gene list of n genes compared against a gene list of m genes when randomly sampled from a universe of u genes:



The lower a similar list p-value, the more significant the overlap between two genes lists, and the less chance there is of random overlap. A high p-value is indicative of how significantly a biological term or pathway is represented in your gene list.