replikation / poreCov

SARS-CoV-2 workflow for nanopore sequence data
https://case-group.github.io/
GNU General Public License v3.0
39 stars 16 forks source link

MinKNOW/Guppy update needs new model for R10.4.1 5 kHz #254

Closed hoelzer closed 1 year ago

hoelzer commented 1 year ago

In a recent MinKNOW/Guppy (v6.5.7) update, the sampling rate for R10.4.1 FCs was increased from 4 kHz to 5 kHz. In Guppy (v6.5.7), there is a new model for 5 kHz: DNA r10.4.1 e8.2 400bps 5khz

The corresponding Medaka v1.8.0 model for 5 kHz data is

r1041_e82_400bps_sup_v4.2.0

In the current ARTIC container we are using, Medaka is v1.7.2. Thus, I created a new ARTIC container w/ Medaka v1.8.0 and the most recent models:

nanozoo/artic:1.3.0-dev--e22cdbc

I suggest, after testing, adding that new container to the pipeline also to allow variant calling with recent Medaka models.

MarieLataretu commented 1 year ago

Thanks, @hoelzer!

The medaka processes fail at the moment, because keras_preprocessing is not found: ModuleNotFoundError: No module named 'keras_preprocessing'

I added the full err file: tmp_medaka.err.txt

hoelzer commented 1 year ago

Hey, thanks for testing @MarieLataretu !

Hm, I did not change anything else besides the Medaka version and everything installed successfully according to conda.

I can try to add a

conda install -c conda-forge keras-preprocessing

to the container build.

MarieLataretu commented 1 year ago

yeah, let's give it a try!

hoelzer commented 1 year ago

@MarieLataretu

docker pull nanozoo/artic:1.3.0-dev--784493e

hoelzer commented 1 year ago

And I deleted already the non-working container nanozoo/artic:1.3.0-dev--e22cdbc

MarieLataretu commented 1 year ago

perfect, nanozoo/artic:1.3.0-dev--784493e worked!