Hi Adam, Great comments. 1.) Nathan's people requested this as well. So yes, I believe we will have a third type called "mutation" where this will be handled. 2.) Media species concentrations collected over time is a very important dataset we need to handle. In the proposed structure, this would be handled in the "Metabolite measures" object, which will be able to handle uptake rates, media concentrations, and intracellular metabolite concentrations collected over time. In fact, I believe we will introduce algorithms that compute average uptake rates based on media concentrations collected over time, as uptake is not actually a directly measureable quantity. In these cases, I think the concentrations listed in the media formulation would be "starting concentrations". Alternatively, it's arguable that we shouldn't put concentrations in the media formulation at all, as these will vary too much. No hard and fast decisions here. Just explaining the thinking behind the proposed structure. Chris From: Adam Paul Arkin [mailto:[email protected]] Sent: Wednesday, April 04, 2012 8:10 PM To: Christopher Henry; 'KBase Microbes List' Subject: Re: [Kbase-devel] Phenotype data format proposal Hi Guys, Looking really good. Two quick questions on a skim- 1) Are you not going to include point mutations/replacements along with insertions/deletions (given the increasing use of MAGE, etc), 2) In the data file info on media would seem to need to be quantitative- so the amount of glucose, etc. should be noted. If there is a time series this might be measured in time. This would seem to be important to the analysis. where will that data be kept? Adam From: Christopher Henry <[email protected]> Date: Wed, 4 Apr 2012 16:49:37 -0500 To: 'KBase Microbes List' <[email protected]>, <[email protected]> Subject: [Kbase-devel] Phenotype data format proposal Hello all, Attached is the proposal for phenotype data in KBase that was discussed in the developers call today. This is geared towards microbes, but probably works for other cases as well. This is a draft, so modification can certainly be made. Also, after discussing this with Paramvir and Ben, it's been decided that we will also introduce a data format for "growth phenotypes", which Ben is in charge of formulating. "Growth Phenotypes" will be highly derived datasets (basically knockouts/fitness pairs). "Growth Phenotypes" will ultimately be computed from the experimental data-structure described in the attached file, and "Growth phenotypes" will be used as direct inputs into FBA and Growmatch algorithms. Chris _______________________________________________ Kbase-devel mailing list [email protected] https://lists.kbase.us/mailman/listinfo/kbase-devel