The taxonomic profile for each sample will be located under the temp/ folder of a HUMAnN run, and it will be available before other files (e.g. genefamilies.tsv) finish. Another option is to run MetaPhlAn on the samples outside of HUMAnN to build your table. You can then pass a pre-existing taxonomic profile for a sample to HUMAnN (so it skips that step) using the --taxonomic-profile flag.
Hi! I just seen this threads. But I am experiencing the same as Silas. I am also working on a cluster managed by SLURM and ran this task three times (2 were resume scripts after the first task hit wall time after 16h). Now its been 2 days and it seems no new files have been generated since saturday. Seems a bit odd, as the logs have not been updated (for reference, 98% of my readings were unclassified by metaphlan), and the last update was that diamond started. I am using 24 CPUs and 62Gb memory.
Thanks!