Pipeline should support long-read submissions end-to-end

Description of feature

The pipeline almost supports long-read submissions but requires two small tweaks.

1. Modify accepted filetype for raw read submission to allow base-called reads in either FastQ or BAM format for PacBio and ONT. The ENA docs also allows tar.gz submission of raw data files, but this might be discontinued in the future. This task also includes speaking to ENA to make sure the documentation is up to date.
2. genome_uploader needs to accept single-contig genomes for long reads, as type chromosome. Complete and deploy the existing PR in the genome_uploader repo.
3. Add required parameters to NF for this step to upload single-contig genomes.
4. Test end-to-end submission of a LR sample. Raw read BAMs can be very heavy and time-consuming to upload, so this needs to fail and try again gracefully.

Start by locating the raw-read filetype handling and the Nextflow parameters for the genome upload step. Review the existing genome_uploader PR and confirm the ENA submission requirements for PacBio and ONT formats. Done means single-contig long-read genomes and FastQ/BAM submissions work end to end, including graceful retry after heavy uploads.