Generate summaries of the tables for the individual runs of qiime dada2 denoise-single. How many features were defined in the first run? How many features were defined in the second run? How do these numbers compare to total number of features after merging?
I get that the first run generates 617 features and 320 210 tot freq and the socond run have 461 features and 155 190 tot freq. The merged table though, have 377 features and 48 925 tot freq.
Why is there a decrease? Is it only fetures existing in both runs that are present in the merged table? I have tried to find information on how this function works in order to resolve this question but havn’t figured it out.
I just crossed-referenced your feature counts, and it looks like you might have mixed files from both the 1% and the 10% datasets while running the tutorial.
From the tutorial:
In this tutorial we’ll work with a small subsample of the complete sequence data so that the commands will run quickly. You can choose either a 1% subsample of the reads or a 10% subsample of the reads. If you’re just trying to gain experience with preparing and combining multiple sequencing runs of data, you can work with the 1% subsample data so that the commands will run very quickly. If you’re using this tutorial to gain additional experience in generating and interpreting QIIME 2 analysis results, you should work with the 10% subsample data so that the results will be supported by more sequence data (1% of the reads is likely not sufficient to support some of the findings of the original study).
The merged 1% data has 377 features (table 1: 263 features, table 2: 177 features). I don’t have a copy of the 10% fully analysed here in front of me, but table 1 for the 10% has 617 features, which leads me to think that some files got mixed up somewhere in there! I would recommend re-running the merge step to make sure you merge the right files, then check the feature table summaries again.