Major batch of changes for eLife revisions

[tomsasani]

Batch of manuscript edits for eLife

1. Implement bootstrap confidence intervals for distance peaks.

   • Rather than using an arbitrary +/- 5Mbp window for defining bounds around AMSD peaks, we now use a bootstrap confidence interval.

2. Remove discussion of Mbd4 as a candidate gene from the Results and Discussion.

   • Using bootstrap confidence intervals, we now exclude Mbd4 as a plausible candidate gene, and instead focus largely on Ogg1 as the most likely gene underlying the chr6 mutator phenotype.

3. Additional simulations to establish AMSD power over QTL mapping.

   • We now perform simulations in which mutation counts are allowed to vary by a factor of 20 across simulated haplotypes. These simulations demonstrate a power advantage for AMSD over QTL mapping.

4. Additional caveats regarding the applicability of AMSD in other populations (i.e., non-RIL populations).

5. More focus on eQTLs (specifically for Ogg1 expression) as possible causal alleles underlying the chr6 phenotype.