MSstats result issue wei zhang  2015-09-07 07:45
 
Hi MSstats team,

Here I met a weird problem for MSstats analysis.

(1) I have three samples, each with three replicates, were set in Skyline.

(2) However, the result showed only single logFC for each protein without indicating which two samples the value calculated from.

(3) the adj. p-values of all of 3,000 proteins were below the threshold.

It is so weird but I cannot figure out where the problem come from and thus need your help. Thanks a lot.

Best regards,

Wei Zhang
 
 
Meena responded:  2015-09-07 09:42
Hi Wei,

1. Document grid
   1) BioReplicate column : if three replicates means technical replicates (from the same sample (same patients or same cell line), just repeat measurements-3 MS runs), it should be the same in each condition. (Here I assume different patients or different mouse per condition) such as for ABA condition, 'BioReplicate=1' for all 3 rows, and for BR condition, 'BioReplicate=2' for all rows. The annotation in Skyline settings.png means that there are 3 condition (ABA, BR, CK) measurements for each patients. Please change them based on your experiment.

2. MSstats group comparison window
   1) Select group(s) to compare against : could you try to use select one group, such as 'BR' only and then 'CK' for next group comparison analysis? if you have more than two groups in 'select group(s) to compare against', it combines the groups being compared in to one big group and give you one fold change. (BR and CK) - ABA
   2) Name of comparison : this should be 'label' column for the result to know which comparison you used. I'm not sure why just random label is shown. Could you try to use any character (letter) label? such as for comparison between ABA and BR, 'BR-ABA' (group in 'select group to compare against:' - group in 'Control group:')

  If you still have the problem, please let us know.

3. It can happen with small samples for expanded scope of biological replicates.

hope to help

-Meena