Add prepare_genomes pipeline to lyrebird_metapackage_creation#270
Add prepare_genomes pipeline to lyrebird_metapackage_creation#270wwood merged 7 commits intowwood:mainfrom
Conversation
|
Hi Rossen, I'm tempted to merge this without looking too hard - it is self-contained and won't harm other code - seem like an OK idea? Is there any things you know are amiss? |
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 9544940f9c
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| genome_to_length[uvig] = int(row['length']) | ||
|
|
||
| new_taxa = {} | ||
| df = pl.DataFrame.read_csv(vcontact_assignments, infer_schema_length=10000) |
There was a problem hiding this comment.
process_vcontact3_taxonomy.py reads the vContact output with pl.DataFrame.read_csv(...), but the project pins Polars 1.x (prepare_genomes/envs/ictv-download.yml), where CSV loading is done via pl.read_csv(...); this causes an AttributeError before any taxonomy rows are processed, so the process_taxonomy rule cannot produce final_reconstructed_metadata.tsv.
Useful? React with 👍 / 👎.
| else: | ||
| last_known = 'unclassified' | ||
| taxbuilder.append(tax_ranks[tax_levels.index(level)] + last_known) | ||
| taxbuilder.append('s__' + (draft[6] if len(draft) > 6 and draft[6] != '' else 'unclassified')) |
There was a problem hiding this comment.
This appends s__ to draft[6], but draft is populated from taxonomy strings that are already rank-prefixed (e.g. ICTV taxonomies are built as d__/.../s__... in ictv-download.py), so genomes with a vContact assignment will be written with malformed species labels like s__s__MySpecies, corrupting downstream taxonomy parsing.
Useful? React with 👍 / 👎.
| done = touch("vcontact3_prepared.done"), | ||
| concat_genomes = "vcontact3_concat_genomes.fna", | ||
| shell: # there are many files, so use cat with find | ||
| "find {input.galah_dir} -name '*.fna' -exec cat {{}} + > {output.concat_genomes}" |
There was a problem hiding this comment.
The same galah_clusters directory is treated as *.fasta in prodigal_gv but as *.fna here; when representatives are emitted as FASTA files, this find matches nothing and produces an empty vcontact3_concat_genomes.fna, causing vContact3 to run on zero genomes (or fail) and invalidating taxonomy reconstruction.
Useful? React with 👍 / 👎.
2 participants
Automated pipeline for most of the front-end preparation for making a lyrebird metapackage (proteins, transcripts, taxonomy)