I have compound concentrations of 30 samples collected from 5 animals (A,B,C,D,E) at six time points (0,1,2,4,8,24 Hours)using LC-MS analysis. Which pattern searching analysis would be best for this kind of data, I would like to compare patterns timewise in all animals keeping in mind respective control condition 0 hr in each animal.
Hi Dipali,
you have no replicates (not ideal) and did not mention for how many lipid species you have (relative or absolute) quantitative data. I assume it is 100++. I further assume you are looking for pattern with respect to your time course.
I would normalize values to the mean of the animal for each lipid and than start with an ANOVA (factor time point) to sort boxplots of your lipids after P-values obtained in this ANOVA. Than look at the first 50 boxplots to get a feeling for which patterns occur. Than decide on a pattern recognition tool.
Bests, Jan
Jan Lisec
Hi, I have replicates (more than 3), over 300 lipids relative quantitative data .
Thank you for suggestions.
I have found a tool in Metaboanalyst for pattern hunting , https://www.nature.com/articles/nprot.2011.319 , representative example of output as attached. They have used mean and average of time variable, and I am particularly interested in paired data comparison.
ex. comparison between
A_0hr vs A_15min, B_0hr vs B_15 min, C-0hr_C15 min and patterns which are consistent in pairs
Regards
Dipali
- Badges
- Science topic
- Similar topics
- Clinical Pharmacology
- Dosing