Dear All
I am doing the metabolomics data analysis. My PCA plot showed tight clustering of the QCs, so i went ahead for picking up the markers. To my surprise some of the ions were matching with m/z of drug molecules. I am in a dilemma whether to remove these ions from the data as they do not represent the markers or keep them in the data. Can anyone suggest me a suitable method to extract the markers from the data. FYI, the biological matrix is human urine.
Regards
Pasha
Dear Mohammed,
Are you sure that the urine originated from subject(s) who has been never taken drug(s) previously?
Well the cohort consists of people who took some drugs..but what drugs no information available.
I think, if you are sure (exact MW, Molecular formula, isotope pattern, fragmentation's) the m/z belong to certain drug(s), you should not use for further analysis (PCA etc). Are the m/z of the certain drug(s) found in certain urine samples, or in all urine samples? If the m/z s were found in all urine samples, I think you should be careful before rejecting the data.
In addition, for confirming the identity of drug(s), it will be nicer if you can compare to the MS data of authentic standard(s).
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Bioinformatics