Make the preprocessing less ukbb specific

[endast]

This PR updates the preprocessing pipeline and makes it more generalized (less UKBB specific). The biggest changes are the following:

• Remove reliance on pvcf_block.txt
  • We now just need vcf_files_list.txt which lists the path to the vcf to be included.
  • We search for files, for example exclude variants recursively
  • The chromosomes flag is now used to filter the included files by reading the first line of the file and getting the chromosome from the file.
• Read tsv files using the pyarrow engine
• Making QC optional
  • The qc part of the preprocessing pipeline now exists in a separate pipeline
  • We have two snakemake pipelines, one with qc and one without qc preprocess_no_qc.snakefile preprocess_with_qc.snakefile

Updated docs for this pr is here: https://deeprvat--39.org.readthedocs.build/en/39/

Without QC

With QC

Related issue: https://github.com/PMBio/deeprvat/issues/34