Hi I am looking for some online courses to analyse data from targeted LC MS. if someone could tell me some good advice of where I could one one please.
if you have targeted MS data (e.g. PRM, MRM, SRM) then Skyline is the program of choice. You will find plenty information and online courses on the Skyline website (https://skyline.ms/project/home/software/Skyline/begin.view) and on youtube. For targeted MS you will not be able to do a PCA as you are searching for specific peptides. However if you meant standard DDA LC MS/MS analysis MaxQuant is a great program and again tutorials are provided on the website (https://www.maxquant.org/) and on youtube.
sorry maybe the way I wrote my question is complicating my real dobut. In my case I quantified some phtochemicals with standars and now I have been thinking it would be a good idea to take a course to learn how to present my results from the quantification in a paper. I know that in most of these results are presented in tables or sometimes in pca but I am not sure if there are other statistical analysis to present these types of results?.
No (free or cheap) courses of which I aware. The best advice is to find a paper doing something similar to what you are doing that presents the data in a way that you find compelling. In terms of statistics, THE issue is convincing others at a glance that of the statistical range on the value(s) in question (i.e., variance or standard deviation about a mean). In terms of presentation quality plotting software, I recommend Kaliedagraph, which isn't all that expensive and has some good statistical graphic layouts. KG is limited to 2D, however. R is free, and does pretty well too, but has a steep learning curve to make presentation quality graphics. The nature of the data will dictate the presentation format you will use. Bar charts, box plots, ven diagrams, x-y-z linear, log-linear, or log-log, scatter plots (all with meaningful error bars), statistical plots. The big issues you face is that almost all the stats tools and graphing software is set up for Gaussian (parametric data). No one ever seems to check that their data is actually parametric,. however, which leads to lots of really bad conclusions. If your data is not parametric, you will need to look at non-parametric methods, and adapting the software to reflect that reality.