Thread by @CancerConnector

(1/n) Excited that final version of work led by @D4N__ studying evolutionary repeatability and collateral sensitivity is out. #mathevo #clecliniccancer nature.com/articles/s4146… @CCLRI @NatureComms @CWRUSOM

(2/n) This collab between @CompSciOxford @CCLRI @CWRUSOM & @MoffittPSOC is an extension of @D4N__ ‘s initial theory paper that proposed ‘steering evolution’ as a method of extending the use of existing drugs to control infection. #AntibioticResistance journals.plos.org/ploscompbiol/a…

(3/n) using the fitness landscape metaphor, we reasoned that instances of collateral sensitivity need not be repeatable, as evolutionary branch points could lead to multiple peaks - which could have different sensitivity profiles.

(4/n) using a mathematical we translated this reasoning into the hypothesis that collateral response SHOULD BE highly variable, and that instances of collateral sensitivity and cross resistance would occur in the same experiments in different *evolutionary replicates* #mathevo

(5/n) to test this hypothesis, we performed 60 evolutionary replicates (12 here) of gradient plate-based evolution of E. coli (DH10B) in cefotaxime and monitored fitness (MIC) and performed targeted and whole genome sequencing. The variability was striking. #AntibioticResistance

(6/n) this variability in primary (cefotaxime) resistance translates into wide variability in secondary sensitivities as well. (We included all data in supplements but only show some of the secondary drugs here) #mathevo #AntibioticResistance

(7/n) to determine statistical significance between evo-replicates MICs @D4N__ implemented a maximum likelihood estimate based method in #python in a #jupyter notebook which you can find here: github.com/Daniel-Nichol/… #openscience

(8/n) because of the (hypothesized and observed) variability in collateral sensitivity, we came up with a new way to quantify how often it occurs, which we term collateral sensitivity likelihood (CSL). #AntibioticResistance #mathevo

@ASMicrobiology @kevinwood18 @NSF_BIO @D4N__ @NIAIDFunding (9/n) while the whole genome sequencing largely validated the information we gleaned from targeted sequencing, several deletions and other small modifications were found by our collaborator Mark Adams at @jacksonlab - full WGS data @NCBI SRA here: ncbi.nlm.nih.gov//bioproject/PR…

(10/n) Take home message - and generalization we hope to make - is that in any study of the evolution of resistance, variability in collateral response is a natural consequence of the process itself. It’s a feature, not a bug. Eager to extend finding ms to #mathonco and #radonc

(11/11) and as an endcap - here’s the link to the paper again, and a picture of me and @D4N__ three days before we thought up the main idea for this paper. #mathmustache #mathonco nature.com/articles/s4146…