Modularize code

[ielis]

Hi @pnrobinson ,the PR aims to split Lirical into smaller pieces, remove the redundant code, and upgrade the dependencies.

Java is bumped to 17, we use Phenol v2.0.0-RC1. The code uses biodownload in the download command.

The CLI includes 3 commands for running analysis, the commands differ by how the input data (path to VCF, phenotype, etc..) is provided:

• phenopacket - input data provided as Phenopacket (only v1 at the moment)
• yaml - input data provided as YAML, and
• prioritize - the entire input data is provided via CLI arguments.

There are many CLI arguments, thus the analysis commands are split into 4 argument groups:

• Input - described above, command dependent
• Resource - paths to lirical data, exomiser data, and background frequency. Will include custom pretest probabliity file in future.
• Configuration - options such as --global, --threshold, --strict. All options have default values here.
• Output- options for writing results, include output folder, prefix, and desired output formats.

This PR is still a draft, I haven't finished the non-analytic commands (simulate, grid, etc.), nor any non-standard (age of onset-related) functionality.

If you have a chance, please try to run a file or two to check if it works OK.

[ielis]

Hi @pnrobinson , I think that this branch amassed a lot of material and we should consider merging to develop. I think we're close to v2 release but I'd like to ask you to look at the code and results to make sure I did not do anything stupid. Can we plan on touching bases in the next days?

[pnrobinson]

@ielis let's go over this tomorrow. In particular, we should run a test suite over the 384 phenopackets and check we are getting the same results!