Specify numpy version prevents error #15

[passelma42]

Pip install numpy==1.19.5 is not soliving the current dependency issue anymore. Would installing medaka with a python env solve this dependency issue?

[MaestSi]

Hi @passelma42, I'm facing the same issue too. Do you know why newer numpy versions are causing the issue? Is that due to the fact that they result in the installation of a newer medaka version, with different syntax?
SM

[passelma42]

Hi @MaestSi,
Not sure. The way I see it, is that the yml file doesn't specify any version for any of the used dependencies... . Installing the pipeline when it just got out VS installing it today is totally different. For instance, Medaka right now is advised to be installed trough pip in a dedicated pip env, with newer versions not supported any longer through conda (if I remember correctly). This being said, there is another tool out there to analyze UMI's which has been working great for us. Maybe it can be customized to help you out as well: https://doi.org/10.1101/2023.06.19.544637.
Greets
Pieter

[MaestSi]

Hi @passelma42 ,
in case you are still interested in the topic, I also wrote a Nextflow implementation of longread_umi pipeline with some adjustments, named UMInator.
Best,
Simone