RefineSchema reimplementation

[rfm-targa]

This PR adds a reimplementation of the RefineSchema/IdentifySpuriousLoci module.

The new implementation aims to simplify and optimize the whole process. This was achieved by reducing the complexity of the workflow, which had multiple intermediate steps that were redundant or unnecessary to efficiently determine the end results, and of the data structures used, which were too nested, affecting readability and the long-term maintainability of the code. The reimplementation aims to improve the accuracy of the results and to provide output files with the most relevant data for users to be able to make well-informed decisions when evaluating the recommendations.

The main changes added by this PR are the following:

• Implemented a more simple approach to evaluate the loci in schemas or a FASTA file with CDSs.
• Removed the majority of the code used by the previous implementation.
• Removed the schema_vs_schema run mode. The schema run mode now accepts any number of schemas.
• Add parameters to define and exclude loci based on the percentage of special classifications in the allelic profiles and the frequency of the loci in the dataset.
• Add an integer identifier to the group separators in the recommendations file to allow searching for specific groups.
• Add the creation of a TSV file with data for all Query-Subject matches evaluated to determine the final recommendations. This file includes data that can help users decide the action to apply for each specific locus/representative CDS in the recommendations file.
• Updated class assignment to better match the logic flow described in the documentation. Added assignment of class 6 for cases that only match through BLASTn, class 7 to cases that have no matches, and class 8 for all loci/representative CDSs that are excluded due to having a high percentage of special classifications and/or being infrequent.