subcommand find/rename

[ammaraziz]

Subcommand for finding fasta files specific to a batch.

Example usage:

flukit find \
    --input-dir {Path} \
    --input-meta {tsv or csv} \ 
    --batch-num {num} \
    --output-dir {Path} \
    --split-by gene

• optional split-by can be split by gene where output is ha.fasta na.fasta etc or all returning all files individually split
• optional input-dir to the directory of fasta files. Encode default paths (see below)
• optional input-meta from fuzee output containing seqno to retreive

Operation:

1. Get a list all fasta files from input directory
2. Parse meta file to extract Seq No
3. Match lists, subset, get complete path
4. Concat files together and write out
5. If split-by is specified, split as appropriate
6. Create folder in output-dir with named batch-number
7. Write out fasta files

Default paths to search in:

• S:/Shared/WHOFLU/mol_biol/00-New Sequences/
• S:/Shared/WHOFLU/mol_biol/Sequencing-NGS/
• /mnt/Sdrive/WHOFLU/mol_biol/00-New Sequences/
• /mnt/Sdrive/WHOFLU/mol_biol/Sequencing-NGS/

Considerations:

• How should NGS runs be handled? These files are located in a different directory and in a single fasta file. Possible solution is to search in /00-New Sequences/ first and if no matches are returned, search in .../Sequencing-NGS/ for any files matching the batch number then split files into .../00-New Sequences/ but do not overwrite existing files.
• Use fuzee api to retreive batch metadata related to #4

[ammaraziz]

For the rename branch to be merged into main, I need to fix or add the following:

• [ ] The controls for how headers are renamed (add_month, add_gene, add_passage) is hidden under a single flag rename. Should these be exposed?
• [x] Handle if the date is missing
• [x] Handle if the designation is missing
• [x] output unmatched records and print this to terminal
• [x] fix the issue with spaces not being replaced with _.
• [x] Handle ascii character ' -> replace with nothing ''
• [x] There's an issue with write_sequences, it expects a dict which works for when sequences are renamed. If they are not renamed, a list is given to write_sequences. Two options, write_sequences detects the object type and acts accordingly or a new function needs to be written.