Set defaults and improve region naming in loaders

[kal26]

Summary

Sets defaults for qc_method, pip_cutoff_to_skip_ind and pip_cutoff_to_skip_sumstat when multiple contexts are used, and regions when only region_name_col is passed or nothing is passed.

Changes

• Setting a real default for qc_method via match.arg("rss_qc"). (This was mentioned in the docs but not implemented)
• Allowing default PIP skip cutoffs to be used with multiple contexts/sumstats. Accept scalar or per-context vector; otherwise error with clear message. If nothing passed, default to 0 for all contexts.
• Assigning phenotype names from region_name_col even without extract_region_name, with a fallback to region_1, region_2, ....

Rationale

Removes the need to always pass verbose arguments for common cases, sets defaults to match was is stated in the docs.
Region names are needed for colocboost_analysis_pipeline to run without errors if multiple regions per phenotype file are analyzed. Previously, names were only assigned if BOTH region_name_col and extract_region_name were passed. This meant in practice you could not pass phenotype files with multiple phenotypes within the region without also passing both these parameters.

Testing

• Checked xQTL-only and xQTL+GWAS run with omitted qc_method
• Checked multiple-context and multi-GWAS cases run with omitted pip_cutoff. Checked that error is throw if pip_cutoff does not match length of phenotypes/susmstats.
• Checked region names are set as expected if only region_name_col is passed, or if no region_name_col is passes and that colocboost_analysis_pipeline runs without the naming error.

[gaow]

Thanks @kal26 from me eyeballing on the codes I don't see obvious concerns of backwards compatibility. The only issue is that we don't have unit tests so it's hard to tell exactly if it breaks anything but @danielnachun i think its fine to merge. Please take a look and do it on your end.

There is one merge conflict possibly due to formatting of white space --- please address that. To avoid white space reformatting, we recommend unifying the R code format before PR,

https://github.com/StatFunGen/pecotmr?tab=readme-ov-file#developers-notes

see the line format_r_code.sh