Arthropod database build errors in version 1.5.1 (Python 3.7)

[JonathanVanHamme]

Hi Jon! Hope you are staying sane and healthy. I'm back to arthropods after a bit of a break, and downloaded the most recent data from BOLD a couple days ago to ensure that the samples we submitted there are included when we do taxonomy assignment here. I noticed you mentioned updating the database build method in v1.4.0, but am not sure if I am getting to the updated instructions:

https://github.com/nextgenusfs/amptk/blob/master/docs/taxonomy.rst

No problems with bold2utax.py and concatenating the resulting files, but the database builds all fail.

I'm running on:
Ubuntu 18.04.3 LTS (GNU/Linux 4.15.0-66-generic x86_64), 24 cores, 378G memory.
In a Conda environment with freshly installed AMPtk 1.5.1 that is working beautifully for fungal and bacterial data.

Looks like all three database builds are calling the same errors (see attached):
2020Nov11_Arthropod_Database_builderrors.txt

[05:05:46 PM]: Now dereplicating sequences (collapsing identical sequences)
Traceback (most recent call last):
File "/results/Miniconda/miniconda3/envs/amptk151/bin/amptk", line 788, in
main()
File "/results/Miniconda/miniconda3/envs/amptk151/bin/amptk", line 779, in main
mod.main(arguments)
File "/results/Miniconda/miniconda3/envs/amptk151/lib/python3.7/site-packages/amptk/extract_region.py", line 614, in main
dereplicate(derep_tmp, OutName, args=args)
File "/results/Miniconda/miniconda3/envs/amptk151/lib/python3.7/site-packages/amptk/extract_region.py", line 58, in dereplicate
if not sequence in seqs:
TypeError: unhashable type: 'dict'

Thanks for any help,
Jon

[nextgenusfs]

Hi Jon -- hope you are well. Sorry about that, looks like a typo when I switched to pyfastx parsing. You can install the master over the top of your conda install, just move into that environment and run:

python -m pip install --no-deps --force git+https://github.com/nextgenusfs/amptk.git

[JonathanVanHamme]

Thanks for the quick response, Jon! Worked like a charm! Database all set up and first data set through. Nice to see that you keep adding functionality to AMPtk, it is my favourite pipeline to work with.

[nextgenusfs]

Great! Let me know if any other issues show up.