|Results of Mann–Whitney U tests for significant differences in periodicity indices between sequences with and without prophagesa|
|Index||Number of tests that yielded p ≤ 5%|
a We used a technique similar to the bootstrap test to reduce the data bias towards model organisms, which often have multiple sequenced strains of the same species. A single genome was randomly selected to represent each species in our dataset while all other genomes of the same species (if more than one genome was available) were excluded from the subsequent analysis. The resulting dataset was divided into two groups based on whether the chromosome contained a prophage or not. The Max2, Max3, and MaxMax indices in both groups were subsequently compared using the Mann–Whitney test. The test was repeated 1000 times with different random selections of representative genomes. The table shows the number of tests that led to rejecting the null hypothesis when using a standard 5% probability cutoff.
Abel and Mrázek
Abel and Mrázek BMC Genomics 2012 13:188 doi:10.1186/1471-2164-13-188