I thought gene flow causes variation in synonymous diversity/divergence.
But this paper shows variation in synonymous diversity/divergence despite the complete absence of gene flow.
To be fair to myself, there's nothing logically wrong with my paper, and the gene flow hypothesis is plausible. It's possible that gene flow does contribute to synonymous genetic variation in E. coli, but I'm not aware of good evidence for this claim. This afternoon, I did a quick analysis of HGT events in research recently published in Nature by Ilana Brito and colleagues. I mapped these events onto the E. coli genome, and while I did find hundreds of HGT events that map onto the core E. coli genome (in contrast to the thousands that map onto the flexible E. coli genome), I found no evidence of a relationship between synonymous diversity and number of HGT events (i.e. strength of gene flow) in the core genome. A lack of evidence doesn't imply that my hypothesis is false, but given the data, the support for my hypothesis is pretty thin at best.
My paper hasn't attracted any citations, probably because my hypothesis is probably wrong! Contrary to my previous beliefs, genomic variation in mutation rate is the most probable cause. Science works, despite the errors made by fallible human beings.