Treatment similarity/specificity statistics

[jeffreybarrick]

Motivation: Many experimental evolution studies compare which genes are being mutated in one or more treatments to understand which ones are driving adaptation to a specific environment.

Implementation: As a post-processing script that is separate from breseq. Use Python or R to calculate similarity statistics and produce relevant graphs. Input is a set of input GenomeDiff files that have #=TREATMENT metadata lines added. The first version should implementing calculations from Figure 2 of Deatherage et al. Specificity of genome evolution in experimental populations of Escherichia coli evolved at different temperatures. Proc. Natl. Acad. Sci. U. S. A. 114: E1904–E1912. [PubMed]. @danieldeatherage can assist by sending his existing code that needs to be adapted.

• Input a set of GenomeDiff files (use relevant Python/R package for reading this information in, don't re-implement this step!)
• Use Biopython/Bioconductor to associate each mutation with a gene using the rules from the publication.
• Replicate shuffling analysis to give p-values for Dice similarity differences between all samples in different treatments (Figure 2A).
• Replicate analysis that predicts which genes are significantly associated with one or more treatments (Figure 2B)
• Apply code to new experiments and come up with expected results and tests.

[jeffreybarrick]

Here are some annotated GD files for the temperature evolution experiment to get started:
https://barricklab.org/release/tmp/TEE-clone-curated-annotated.tgz