Cover Image

HARDBACK
$69.00



View/Hide Left Panel

matrix R2 between the combined gene set of the proteasome and NPC for each species. We optimized the modularity quality function to partition this combined matrix into groups in a data-driven manner. We next asked whether this data-driven partition was statistically similar to the true partition of the genes into the two groups of proteasome genes and NPC genes. To answer this question, we computed the partition similarity between the data-driven partition and the true partition and used permutation testing to determine whether this similarity was statistically significant. The permutation test was implemented by randomly reassigning genes to the two groups of “proteasome” and “NPC,” recomputing the correlation matrix R2ʹ, partitioning the genes in the correlation matrix into modules, and computing the similarity between this partition and the true partition. This process was repeated 1,000 times to construct a distribution of similarity values expected under the null hypothesis that the coregulation patterns between proteasome and NPC genes do not differ. For each species, the P value to reject this null hypothesis was computed as follows: the number of similarity values derived from the permuted data that were greater than the real similarity value, divided by the number of permutations.

ACKNOWLEDGMENTS

We thank Boris Shraiman, Mason Porter, Adel Dayarian, and Marija Vucelja for invaluable suggestions and comments; Scott Grafton and Jean Carlson for facilitating the collaboration; and Marc Kirschner and Leonid Peshkin for sharing Xenopus tropicalis microarray data. This work was supported by gifts from Harvey Karp and Gus Gurley (K.S.K.); David and Lucile Packard Foundation (D.S.B.); Public Health Service Grant NS44393 (to D.S.B.); Institute for Collaborative Biotechnologies Contract W911NF-09-D-0001 from the U.S. Army Research Office (to D.S.B.); and the Australian Research Council (S.M.D. and B.M.D.).

The data reported in this chapter have been deposited in the Gene Expression Omnibus (GEO) database, www.ncbi.nlm.nih.gov/geo (Accession No. GSE29978).



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement