search

broadinstitute / protigy

Proteomics Toolset for Integrative Data Analysis
Other
22 stars 16 forks source link

Analysis with only selected columns #11

Closed anananwal closed 2 years ago

anananwal commented 2 years ago

Since the recent update, which includes the option to chose Experiment or group column for testing, I can't seem to do statistical tests on only a few selected replicates (to leave out dropouts for example). If I do not fill in the experiment column in the experimental design file it still includes this sample in the analysis. Previously leaving the experimental column cell blank for a sample excluded it from analysis. Is there a new way of doing this?

Example: When trying to exclude one replicate from one experiment it puts that sample as an unnamed experiment with one sample in (see first attached screenshot). Screenshot 2022-07-29 at 12 27 40 Then I get this error when I try to run a moderated t test: Warning: Error in apply: dim(X) must have a positive length

Example 2: See 2nd attached screenshot of 8 unwanted samples being included together as unnamed experiment.

Screenshot 2022-07-29 at 12 17 43 Screenshot 2022-07-29 at 12 27 40
nmclark2 commented 2 years ago

Hi,

Samples that you wish to exclude in the experimental design file must use NA instead of "" now. This was changed due to an issue in how Protigy interprets blank values. If you re-download the experimental design template from Protigy, you will notice that the default value is now NA instead of "" Changing blank values to NA in your experimental design file should fix this issue. Please let me know if it does not.

Thanks, Natalie

nmclark2 commented 2 years ago

@anananwal were you able to resolve this issue? Please let me know if not. Otherwise I will close it. Thank you.

anananwal commented 2 years ago

Yes, I did. Thank you so much for the quick solution. Sorry I didn’t reply!

On 19. Sep 2022, at 22:08, Natalie Clark @.***> wrote:



@anananwalhttps://github.com/anananwal were you able to resolve this issue? Please let me know if not. Otherwise I will close it. Thank you.

— Reply to this email directly, view it on GitHubhttps://github.com/broadinstitute/protigy/issues/11#issuecomment-1251497414, or unsubscribehttps://github.com/notifications/unsubscribe-auth/A2HAJZ6EZEVSUQWEXNRBVITV7DB4PANCNFSM55AJE7TQ. You are receiving this because you were mentioned.Message ID: @.***>