Hi @Ishanmanandhar,
So, the thing you specify for --m-metadata-column needs to be a column in your metadata file. The “meaning” of subject, or of any other metadata column name, is dependent on your specific metadata file. In the case of the Moving Pictures tutorial data that the q2-aldex2 tutorial uses, subject refers to the individual who provided a sample (subject-1 or subject-2).
The --m-metadata-column for q2-aldex2 defines “the different groups you will be testing”. If you’re going to use q2-aldex2 to look at differentially abundant features between – in your case – IBD and “control” samples – you will need to figure out what metadata column indicates whether a sample is “IBD” or “control” (or something else? you may want to filter the table first, depending on the question you’re asking), and then use this column as --m-metadata-column when you run q2-aldex2. You’ll need to figure this out regardless of what differential abundance method you use, including ANCOM, Songbird, gneiss, etc.