Chimera checking should be modeled after mothur where you can break the analysis up by sample (since chimeras form within the reaction). This might be best done before OTU calling - but that would be a trade-off between speed and quality. It will be easier to identify chimeras with all of the data present before clustering, but it might take way longer and that different at the end from doing it straight.
The results of chimera checking should be a list of chimeras, which might be as easy as throwing the flag to show chimeras at usearch.
Chimera checking should be modeled after mothur where you can break the analysis up by sample (since chimeras form within the reaction). This might be best done before OTU calling - but that would be a trade-off between speed and quality. It will be easier to identify chimeras with all of the data present before clustering, but it might take way longer and that different at the end from doing it straight.
The results of chimera checking should be a list of chimeras, which might be as easy as throwing the flag to show chimeras at usearch.