Add --output-matched-protein-sequences option to supplement mode

[wwood]

Motivation

• Provide an option to export protein sequences that matched HMM hits during supplement so users can inspect matched proteins alongside the matched transcripts.
• Keep the same read renaming convention using the ‡ delimiter so sequence identifiers remain compatible with downstream code.

Description

• Added a new CLI argument --output-matched-protein-sequences in singlem/main.py and passed it into the supplement workflow as output_matched_protein_sequences.
• Propagated the argument through Supplementor.supplement, updated generate_new_metapackage and gather_hmmsearch_results signatures, and passed the value into worker calls.
• Implemented writing of matched protein sequences inside run_hmmsearch_on_one_genome by appending protein FASTA entries for transcripts in matched_transcript_ids and pre-creating/clearing the output file in gather_hmmsearch_results when requested.

Testing

• Ran the full test suite with pixi run -e dev pytest test as required by project docs.
• Test run collected 244 items and completed with 2 failed, 222 passed, 20 skipped, 1 warning.
• The two failing tests are test/test_pipe.py::Tests::test_read_chunk_size_forward_gzip and test/test_pipe.py::Tests::test_read_chunk_size_paired, which failed due to a --read-chunk-size must be divisible by 4 validation error.
• All tests unrelated to supplement changes passed.

---

Codex Task