Private Klebsiella-only plasmid-vs-host training dataset. Plasmid segments were sampled from NR Klebsiella plasmid representatives, host candidate segments were checked against the all-NR plasmid minimap2 reference, and the clean training config is balanced 1:1 plasmid:host by fragment count.
This is not the all-NR plasmid genus-balanced dataset and not the older NR-cluster curriculum dataset.
Generated: 2026-06-05 11:14:40 CEST… See the full description on the dataset page:
https://huggingface.co/datasets/neuralbioinfo/klebsiella-plasmid-training.
