Error while running minpath, one sample only

Microbial community profiling HUMAnN
1

Dear HUMAnN creators,

I’ve run humann3 successfully on loads of samples, with a single exception. For that particular sample, I get the genefamilies, minpath starts running, but then there is an error (see below).

I ran the same sample without any issues with the humann2.

This is the error message:

Traceback (most recent call last):
File “/miniconda3/envs/metagenomes/bin/humann3”, line 11, in
sys.exit(main())
File “/miniconda3/envs/metagenomes/lib/python3.7/site-packages/humann/humann.py”, line 1119, in main
gene_scores, reactions_database, pathways_database)
File “/miniconda3/envs/metagenomes/lib/python3.7/site-packages/humann/quantify/modules.py”, line 154, in identify_reactions_and_pathways
utilities.command_threading(config.threads,minpath_commands)
File “/miniconda3/envs/metagenomes/lib/python3.7/site-packages/humann/utilities.py”, line 607, in command_threading
worker.start()
File “/miniconda3/envs/metagenomes/lib/python3.7/threading.py”, line 852, in start
_start_new_thread(self._bootstrap, ())
RuntimeError: can’t start new thread

And the end of the log:

12/31/2020 01:13:45 AM - humann.quantify.modules - INFO: Run MinPath on g__Actinobaculum.s__Actinobaculum_sp_oral_taxon_183
12/31/2020 01:13:45 AM - humann.quantify.modules - INFO: Compute reaction scores for bug: unclassified
12/31/2020 01:13:47 AM - humann.quantify.modules - INFO: Run MinPath on unclassified
12/31/2020 01:13:47 AM - humann.quantify.modules - INFO: Compute reaction scores for bug: all
12/31/2020 01:13:49 AM - humann.quantify.modules - INFO: Run MinPath on all

Thanks a lot!

G

2

Hello, Thank you for your detailed post! It is possible other processes are running on the same machine and it is a bit overloaded? It looks like from the error messages that the software can’t start a new thread. Can you try running with a single thread or check to confirm that the machine is not overloaded and this is not a resource based error? Hopefully one of those two should resolve your issue.

Thank you,
Lauren

3

Thanks, Lauren, for your reply.
It is a cluster, and not an overloaded one. I tried running it with more threads as well as with a single thread, always the same error. I think (but I’m not sure anymore, it was before Christmas) that I also tried to run it on a desktop with the same result.

I’m a bit short on time (surprisingly :)), so I haven’t had time to look if there is a way to try only the Minpath step on the humann genefamilies output without rerunning the whole humann? Somehow I didin’t get the resume option to work as I expected it to.

However, I looked a bit at the sample and the outputs, there seems to be nothing particular about this one, it is not larger or something…

4

In the end it worked on a desktop, with genefamilies as input, with a single thread. Thanks!

5

Great! Glad to hear you got it to work!