Add new experiment processing/ reprocessing rules to gxa

[pmb59]

The aim of this PR is to add new rules and make of use existing ones (recalculations) to address processing of new experiments and reprocessing steps, in order to achieve a super-set of data analysis rules for GXA. This workflow now deals with previous atlas_prod scripts used before moving our systems to new EBI infrastructure.

The super-set of rules can now handle the following experiment types:

• baseline
• differential rnaseq
• differential microarray

This PR includes:

• logic that allows Snakemake rules collecting I/O files depending on the analysis goal (recalculations or reprocessing).
• (re)processing/recalculations only for selected specie/s
• updated differential expression strategy
• a number of Atlas scripts from production (which will be moved to atlas-analysis and then included as submodule here)
• communication to ISL db (only for new experiment or reprocessing)
• Major code changes/ new files are at:

- Snakefile
- Snakefile-reprocess
- Snakefile-sorting-hat
- bin/reprocessing_routines.sh
- bin/round_log2_fold_changes.R (reviewed)
- envs/isl-db.yaml
- run_sorting_hat_test_data.sh

Future merge requests will address:

• replacing generic bash routines with conda atlas-bash-util
• refactoring scripts using a submodule atlas-analysis
• adding proteomics and differential proteomics analysis rules
• add conditional structures that handle cases of optional/required parameters and files