AlexGM

Hi Lepton, Sorry for the mistake, the paper is not out yet, i forgot to remove the sentence in the online version. I removed it and will put it back (and add the link to the paper) as soon as the paper is out. Yes the principle is the same as in Falcor (Hall et al 2009) In practice, the main differences are: - it can take into consideration the differential growth rate of your mutants. If you don't know the fitness of the mutants, bz-rates can estimate it directly from the fluctuation data. As an illustration: in the lab we have mutants with significant fitness improvement and high mutation rates. The combination of these 2 parameters has huge effects on the mutation rates that we compute. - it can correct the mutation rate for the plating efficiency (proportion each culture that is plated) directly. It can be useful if you want to screen several phenotypes or for high mutation rates. - you don't need to run a java code in the browser (which is IMHO painful with modern browsers) because the jobs do not run locally on you computer but on our server. Alex

Hi all, We have set up a web-tool to compute mutation rates from fluctuation assays: bz-rates. This tool implements a generalized version of the Ma-Sandri-Sarkar maximum likelihood method that can take into account the relative differential growth rate of mutant and wild type cells. When this parameter is unknown, you can use a generating function estimator to estimate both the mutation rate (nearly as precisely as the MSS method) and the differential growth rate. The source code is also available if you want to re-use the code or make pull requests I hope it'll be useful, check it out ! Best, Alex