by
lmh » Tue May 08, 2018 9:43 pm
so much depends on your environment. In an academic rather than industry environment, we've got a customer-base who expect to be able to move their data into R, Matlab, proprietary 'omics software, not to mention getting quantitative data into whatever their favourite graphing package happens to be - or possibly trying to make the whole thing visible on a web interface. Excel is just one of many calculating tools they'll use along the way. There's no way that a CDS will ever offer all the functionality they require. Waters and others have tried it, and create very nice pipelines, but you always get someone who remembers that some part of someone else's pipeline does their particular job better. Also, we can't afford to tie ourselves to a single manufacturer, so we don't really want to commit ourselves to a software solution that only works for one family of instruments.
But even when our users just want to do the stuff that Empower and other packages can do, you have to look at the likely sources of error. An experienced Excel user with only a dabbling interest in chromatography is less likely to mess up a calculation in Excel than in Empower custom calculations, and his colleagues are more likely to spot the problem in a workbook that contains the calculations. I encourage those who use my lab to make use of the various systems' abilities to create calibration curves and carry out quantification, but I don't think it would be time-efficient for them to do much more in the CDS. Also, if their QC checks, and sample dilutions, are embedded in software that we perhaps abandon in a few years when we change instruments, no one can decode later what they did (yes, we can keep the old software, but then Microsoft helpfully discontinue support on WinXP or whatever it is, and we find we can't actually run the software we kept, to open the data we kept, so in effect, our archive is gone). Excel may be universally distrusted (even the R people loathe it), but it's been around for a while, and it doesn't look like it's vanishing soon. The R people would far rather us get into comma-separated-values as soon as possible, and stay there!
For what it's worth, I've still got a slide-rule under the desk...