qiita
qiita copied to clipboard
qiita-spots
Would be nice to be able to merge metadata columns when doing an analysis
I remember there being functionality in the past (I think in the old QIIME database) where one could merge info in two different metadata categories "on the fly" when doing an analysis. I thought that this functionality was pretty convenient/useful.
Agree but I think this would be nicer on the visualization per se. For example, within Emperor. As predefining the categories that you want to merge means that you already know what you are looking for, which is not generally the case. What do you think @clozupone?
Yes, I agree - I think adding to emperor would be better. I can add as a feature request there instead…. Thanks! Cathy
On Aug 26, 2015, at 11:10 AM, Antonio Gonzalez <[email protected]mailto:[email protected]> wrote:
Agree but I think this would be nicer on the visualization per se. For example, within Emperor. As predefining the categories that you want to merge means that you already know what you are looking for, which is not generally the case. What do you think @clozuponehttps://github.com/clozupone?
— Reply to this email directly or view it on GitHubhttps://github.com/biocore/qiita/issues/1446#issuecomment-135110666.
If adding to emperor, it might be worth considering how to export the new categories from emperor so that they can be used for all the other analyses in qiime/qiita after they’ve been manually refined?
On Aug 26, 2015, at 10:16 AM, clozupone [email protected] wrote:
Yes, I agree - I think adding to emperor would be better. I can add as a feature request there instead…. Thanks! Cathy
On Aug 26, 2015, at 11:10 AM, Antonio Gonzalez <[email protected]mailto:[email protected]> wrote:
Agree but I think this would be nicer on the visualization per se. For example, within Emperor. As predefining the categories that you want to merge means that you already know what you are looking for, which is not generally the case. What do you think @clozuponehttps://github.com/clozupone?
— Reply to this email directly or view it on GitHubhttps://github.com/biocore/qiita/issues/1446#issuecomment-135110666.
— Reply to this email directly or view it on GitHub https://github.com/biocore/qiita/issues/1446#issuecomment-135112024.
I'm not sure we want to store the merged column in the metadata DB as it will create duplicated fields. However, we could find a way to save those combinations in the analysis. One way that comes to mind is the idea of saving Emperor status via the URL, which you then could save in your analysis and put a description. Then you could share it with other users, as described here: https://github.com/biocore/qiita/issues/1444. What do you think?
@antgonza we might be able to put that column back in the mapping file for the analysis, but do not modify the templates of the actual studies, as those columns might be the result of merging 2 columns from 2 different studies.
I agree, this feature would fit great in emperor and being able to export newly created mapping files will also be tremendously helpful! I also think that Qiita should be capable of creating new columns in the mapping file mainly to make it easier to extend the usability of other tools/scripts that are not as interactive, think of statistical tests, taxa summaries, etc.
@ElDeveloper This is the third or fourth issue I've found related to exporting metadata from Emperor, can we do a status update on whether that idea was abandoned for lack of time or lack of merit?
Right, and for this the main thing that Emperor would need is an endpoint in the server that it can ping when such an action is triggered. We also need a way to select samples in a lasso-kinda fashion, which happens to be kinda complicated in WebGL.
On (Oct-11-17|12:31), Antonio Gonzalez wrote:
Emperor is still in development (Q2 version), slowly but still moving ...
-- You are receiving this because you were mentioned. Reply to this email directly or view it on GitHub: https://github.com/biocore/qiita/issues/1446#issuecomment-335793931
@adswafford & @ElDeveloper, should we move this issues to https://github.com/biocore/emperor so they are not buried in here ...
+1
On (Sep-21-18|14:49), Antonio Gonzalez wrote:
@adswafford & @ElDeveloper, should we move this issues to https://github.com/biocore/emperor so they are not buried in here ...
-- You are receiving this because you were mentioned. Reply to this email directly or view it on GitHub: https://github.com/biocore/qiita/issues/1446#issuecomment-423558596