Refine clustering about swarm HOT 5 CLOSED

One of the files you are calling (swarms, fasta, swarm binary) is missing. The most likely option is that the script does not find the swarm binary: the default is to search for /usr/bin/swarm. If your binary is elsewhere, you will end up with that error message. Use -b /PATH/ to indicate where to find the swarm binary.

from swarm.

Thank you - that worked! I have another question regarding how you normally would use SWARM when wokring with multiple datasets desired to cluster together and study the swarm / OTU distribution for like with an OTU table. When I dereplicate my data as suggested, is there any way to recover the read abundance for each dataset, combine that with the swarm output file and thus produce an OTU table?

I tried with the output using the UCLUST format and Rob Edgar's script uc2otu.py but I noticed that sample identifiers ("barcode=XXX") are removed during replication so it did not work. In addition, the swarm_breaker script does not offer UCLUST formatted output anyway, as far as I understand. Is there any other way to do this, or should I work out an alternative way to dereplicate and write a script for combining dereplication output and a swarm output file to write an OTU table?

Best wishes,
Anders

from swarm.

Thanks for your input. I added a checking function to the python script to produce more informative messages when the swarm binary or the input files are not found or not readable.

Regarding your question on the OTU contingency table, we started a wiki section to present code examples and workflow suggestions.

from swarm.

Do you have any advise for how to produce "OTU tables" like I asked?

from swarm.

Yes. I added a wiki page on that topic: Working with several samples. I close that issue. Please feel free to open a new issue on the "OTU table" topic if need be.

from swarm.