PR for version 0.9.20

Summary

Updated pre-processing for Silva 138.1 (with species)
Updated pre-processing for Greengenes2
Updated rdp_classifier version from 2.5 to 2.14
Optimized lineage2taxTrain.py
Added taxa formatting when no pre-processing for Silva 138.1, Greengenes2, ezBiocloud and UNITE
- Solved convergent taxa in Silva 138.1 and Greengenes2

Details

5627a03e705a4b94b92ed98f8e8152d43a5348e8 Added conda env and container for pandas
d1d38607f51692200602eb55aae455c63ba36232 Added python script for derep_and_merge_taxonomy:
- Fills empty taxa with placeholder names (ex: Escherichia-shigella_s)
- Replaces Incertae, unkown and endosymbionts with NaN (caused convergent taxons for rdp)
- For unclassified taxa, each taxa corresponding to a different cluster will have a unique index
cc8540a Added db_version parameter
6f30c51 Added script to clean taxonomy files when no processing
d574275 Added list of classifiers for preparing databases (RDP, QIIME, and DADA2 by default)
1619919 Added RDP memory as parameter

4a9408188435df1e3c2a146e8dece978448a79b0 R scripts linting
bdb8a0e20af0541dbc465854d67f0c20fe812305 Changed R to python script for derep and merge taxonomy
c79db5a a467bc7 Updated python env and container to 3.12.3
991605e Replaced numbers_species and numbers_genus by a dict {'Species': 2, 'Genus': 4}
9920ea4 Use rdp_classifier instead of rdptools
7c32765 Updated tax_formatting.py to work with Greengenes2

d1d3860 for #42
145cfd0 Fixed lineage2taxTrain.py slowness by replacing list to set (thanks @tpillone !)
35b648c Fixed some convergent taxa by propagating NaN to downstream ranks