PPV analysis

[katholt]

Created function solo_ppv_analysis (in file solo_ppv.R)

• works with input geno & pheno tables in standard format output by import_ncbi_ast and parse_amrfp parsers
• uses getBinMat to create contingency tables for specified drug and class/es
• finds solo markers
• creates solo stats table (x, N, p, 95% CI - for resistance (R vs I/S) and nonwildtype (R/I vs S)
• creates plots showing distribution of S/I/R phenotypes, and PPV (with 95% CI), for each solo marker

This replaces AMRgen_PPV_plots_v5.R which can now be archived

Notes for development

• uses getBinMat, which currently defines nonwildtype as R/I vs S, need to accommodate alternative definitions e.g. from a user-specified column or as.sir interpretations using ECOFF
• would be helpful to add smart features like auto-detecting the right columns to use as drug_agent, drug_class, antibiotic, sir etc instead of assuming or asking users for colnames (same as getBinMat)
• might be helpful to pull out some steps into functions, that could be access directly without running the whole analysis, if we find they are useful in other analysis/plotting functions

[katholt]

updated getBinMat to include the SIR calls in the output (column named pheno, class sir)

so can now update solo_ppv_stats to use this column rather than adding it to the output of getBinMat - DONE