Inferring directed-graph patterns of gene responses to treatments
The analysis of patterns of gene expression to multiple treatments can be challenging with small sample sizes, because certain responses cannot be statistically ascertained. When response patterns are represented as directed graphs, however, additional information about true response patterns might be inferred. We exploited the relationship between sample size and a graph property known as contractibility, and used synthetic replicates to make inference about patterns of gene response. With microarray data from rats' liver cells, we showed that this approach uncovered subtle interactions in response patterns and resulted in more and better functionally enriched gene clusters.
5th International Conference on Bioinformatics and Computational Biology 2013, BICoB 2013
Phan, V., Vo, N., & Sutter, T. (2013). Inferring directed-graph patterns of gene responses to treatments. 5th International Conference on Bioinformatics and Computational Biology 2013, BICoB 2013, 7-12. Retrieved from https://digitalcommons.memphis.edu/facpubs/2900