hyejong/genomeocean-main-classifier

GenomeOcean Main Classifier

A GenomeOcean 100M v1.2 fine-tuned sequence classifier that splits genomic FASTA contigs into three broad biological categories:

Label	Meaning
Cellular	Sequences derived from cellular organisms (eukaryotic, bacterial, archaeal, mitochondrial, plastid)
NCLDV/Mirus	Candidate Nucleocytoviricota (giant virus) or Mirus sequences
Other Viruses	Phages and other viral sequences outside NCLDV/Mirus

This model is the first stage of a two-stage hierarchical Giant Virus classifier. Contigs predicted as NCLDV/Mirus here are meant to be passed to the companion hyejong/genomeocean-sub-classifier model, which further splits them into NCLDV and Mirus. Full pipeline code, preprocessing details, and CLI tooling are available at Genomeocean_Giant_Virus_Classifier on GitHub.

Model architecture

• Base: GenomeOcean 100M v1.2, a Mistral-architecture genomic language model, fine-tuned for sequence classification (MistralForSequenceClassification).
• Hidden size 768, 12 layers, 8 attention heads, vocab size 4,096, max position embeddings 32,768.
• Custom modeling code (modeling_mistral.py, configuration_mistral.py) is included in each fold folder, so loading requires trust_remote_code=True.

5-fold ensemble

This repository hosts five independently fine-tuned folds (fold1 – fold5), each in its own subfolder with a full set of model, tokenizer, and config files. The recommended way to use this model is to run all five folds on the same input and average (soft-vote) the softmax probabilities — this is what the reference pipeline below does automatically. A single fold can also be used on its own, at the cost of losing the variance/agreement signal the ensemble provides.

Input format

Inputs are 5,000 bp genomic chunks derived from FASTA contigs, preprocessed as follows before tokenization:

1. Convert the sequence to uppercase.
2. Remove characters other than A/C/G/T/N.
3. Remove N.
4. Split the cleaned sequence into 5,000 bp windows with a 5,000 bp stride (non-overlapping).
5. Contigs shorter than 5,000 bp, and the incomplete tail of longer contigs, are not used for prediction.

Tokenization uses the fold's own tokenizer with max_length=1250.

Usage

Recommended: reference CLI package

The GitHub repository ships an installable CLI (genomeocean-main) that handles FASTA parsing, chunking, batching, 5-fold ensembling, and result aggregation for you:

git clone https://github.com/hyej0ng/Genomeocean_Giant_Virus_Classifier.git
cd Genomeocean_Giant_Virus_Classifier
python -m pip install -r requirements.txt
python -m pip install ./genomeocean-main-classifier

genomeocean-main predict \
  --input /path/to/input.fasta \
  --output-dir /path/to/results \
  --model-id hyejong/genomeocean-main-classifier \
  --subfolder fold1 --subfolder fold2 --subfolder fold3 \
  --subfolder fold4 --subfolder fold5 \
  --device cuda

See the Main Classifier README for the full CLI reference, or the Integrated Pipeline README to run Main and Sub together and get final Cellular / NCLDV / Mirus / Other Viruses labels in one command.

Direct transformers usage (single fold)

import torch
from transformers import AutoModelForSequenceClassification, AutoTokenizer

model_id = "hyejong/genomeocean-main-classifier"
fold = "fold1"

tokenizer = AutoTokenizer.from_pretrained(model_id, subfolder=fold, trust_remote_code=True)
model = AutoModelForSequenceClassification.from_pretrained(
    model_id, subfolder=fold, trust_remote_code=True
)
model.eval()

id2label = {0: "Cellular", 1: "NCLDV/Mirus", 2: "Other Viruses"}

sequence = "ACGT..."  # a single, already-preprocessed 5,000 bp chunk
inputs = tokenizer(sequence, return_tensors="pt", truncation=True, max_length=1250)
with torch.inference_mode():
    probs = torch.softmax(model(**inputs).logits, dim=-1)[0]

predicted = id2label[int(probs.argmax())]
print(predicted, probs.tolist())

To reproduce the full 5-fold ensemble manually, repeat the above for fold1 – fold5 and average the resulting probability vectors before taking the argmax.

Training data

Trained on metagenome-derived genomic contigs/fragments:

Label	Scope
Cellular	Eukaryotic, bacterial, archaeal, mitochondrial, and plastid sequences
NCLDV/Mirus	Full NCLDV and full Mirus sequences
Other Viruses	Phages and other viruses

Limitations

• Confidence scores are hierarchical/comparative, not calibrated probabilities of biological truth.
• Sequence length, assembly quality, and distance from the training distribution can affect predictions; ensemble agreement across folds (ensemble_agreement, confidence_std in the reference pipeline output) should be checked for low-confidence or borderline calls.
• For final NCLDV vs. Mirus resolution, pair this model with hyejong/genomeocean-sub-classifier.

License

MIT