phylotypr 0.2.0
New features
- Add new function
classify_sequences(), which has confidence score filtering built-in as an option and is multithreaded by default. Can be run by giving the function the name of a data frame and the columns corresponding to the sequence id and sequences or as a strollur object and function will be output results as appropriate
Improvements
- Update vignettes and documentation for new functions and features
- Core functions can now process more than one sequence at a time
- Added type checking and check for column names in
classify_sequences(),filter_taxonomy(),print_taxonomy() -
build_kmer_database()now includesreference_dataargument - modified
filter_taxonomy()to take a data frame rather than a list - modified
print_taxonomy()to take a data frame rather than a list and then output mothur-like taxonomy string to user specified depth. If specified depth is shorter than observed, string gets truncated. If depth is longer than specified it adds “_unclassified”. User can turn off confidence scores -
rdp_trainset9_pdsandrdp_trainset9now include reference data as a data frame in the reference_data slot; will be incorporated in databases provided by{phylotyprrefdata}. These were renamed from trainset9_rdp and trainset9_pds - Changed
idtosequence_nameacross functions - Modify read_fasta to convert non-ATGCU-. to N; force capitalization. This was also applied to built in reference data
- Remove use of
set.seed()to be within call toclassify_sequences() - Optimized
build_kmer_database()andclassify_sequences()for performance making both significantly faster
phylotypr 0.1.1
CRAN release: 2025-02-21
This is a patch release to… * update the citation * change the license to GPL v3 * make a small change to the behavior of the code for when a k-mer shows up more than once in a sequence
