From owner-metabolic-reg@net.bio.net Fri Oct 08 23:00:00 1993 Path: biosci!daresbury!doc.ic.ac.uk!uknet!festival!castle.ed.ac.uk!eang62 From: eang62@castle.ed.ac.uk (Herbert Sauro) Newsgroups: bionet.metabolic-reg Subject: Re: Substance rather than form Message-ID: Date: 9 Oct 93 10:56:09 GMT Sender: news@festival.ed.ac.uk (remote news read deamon) Organization: Edinburgh University Lines: 49 This was sent some time ago but never appeared, sohere's another attempt at sending it: ----------------------------------------------------------------------------- John says: >Everybody appears to define an elasticity as a partial derivative, but >opinions seem more divided as to whether the control coefficient >definitions are best expressed as total or partial derivatives. > >Your comments will be gratefully received ... > > ... John Woods I had a recent argument with one of our distinguished members a few weeks ago on this point who wanted control coefficients to be partial. I almost always use totals for control coefficents, I don't think I've ever used partials. Here is my reasoning...... The reason why I use partials for elasticities and totals for control coefficients is simply to make clear that elasticities refer to a local response, i.e a partial response - only one metabolite is allowed to move - and control coefficients are measured while allowing all metabolites (except the boundaries of course) to move, a total response. Thus with respect to elasticities, we measure only a partial response of an intact pathway, while with control coefficients we measure the total response of an intact pathway. There will be some of course who will say, yes but with a control coefficient we only change the activity of a single enzyme (lets say) and keep all other enzyme activiities fixed, so isn't the control coefficient measuring a partial response aswell? Well yes it is, but we know what the TOTAL response of a pathway is, it's given by the summation theorems. I guess it's also to do with what level we're at. If we're interested in the interaction of a single enzyme with the intact pathway then we can stick to partials for elasticities and totals for control coefficients. On the other hand if we're interested in the interaction of intact pathways with the genetic pathways (which act at more than one enzyme at a time) then maybe we might make the control coefficients partial and the activity changes at the genes (eg promoter efficiency) total. If one ever attempts to derive from first principles the relationship between the control coefficients and elasticites then depending on your organisational skills, it becomes critical to distinguishing between the partials and the total. Herbert Sauro .