phanerozoic / dna-origin-classifier

dna-origin-classifier

A reference-free, alignment-free classifier that labels a DNA coding sequence by its source: human, other eukaryote, bacterial, viral, or engineered/synthetic. It uses no alignment and no sequence database. One fixed k-mer featurizer feeds three linear heads.

Method

• Featurizer. For k = 4, 6, and 8, count k-mers folded onto their reverse-complement canonical form, normalize each k to within-sequence frequencies, concatenate, and divide by a stored per-feature scale. This gives a 35,112-dimensional multi-order composition vector that is exactly strand-symmetric.
• Heads. Three linear readouts on that vector:
  • origin — 5-class head (human, eukaryote, bacteria, virus, engineered).
  • host — binary head, human vs non-host (bacteria/virus).
  • engineered — binary head, engineered vs natural.

All weights live in model.safetensors ( feature_scale , origin.weight/bias , host.weight/bias , engineered.weight/bias ), 280,903 parameters, about 1.1 MB.

Usage

from model import DnaOriginClassifier
clf = DnaOriginClassifier("model.safetensors")

seq = "ATGGCTAGCAAAGGAGAAGAACTTTTCACTGGAGTTGTCCCAATTCTTGTTGAATTAGATGGTGATGTT"
clf.classify(seq)          # -> 'human' | 'eukaryote' | 'bacteria' | 'virus' | 'engineered'
clf.host_score(seq)        # higher = more human/host-like
clf.engineered_score(seq)  # higher = more likely engineered/synthetic

Requires only numpy and safetensors .

Evaluation

Measured from the published weights on the test and novel-taxa splits of dna-origin-benchmark :

Five-class origin accuracy on the held-out test split: 0.65 (random baseline 0.20). The human/eukaryote boundary is the hardest case and accounts for most of the five-class error; the binary host head is the right tool when the question is human versus non-host.

Calibration

The heads output linear margins, not calibrated probabilities. Use the argmax of logits for origin and the ranking or sign of host_score / engineered_score for the binary tasks; do not read the raw values as probabilities.

Training data

All heads are fit on the train split of dna-origin-benchmark . Sequences are RefSeq coding sequences and NCBI synthetic-construct records.

References

• Benchmark, splits, and baselines: phanerozoic/dna-origin-benchmark .
• Comparison model: HuggingFaceBio/Carbon-8B , an 8B-parameter genomic language model evaluated zero-shot on the same splits. This classifier shares no weights or outputs with it and is not a derivative.
• Method lineage: k-mer naive-Bayes sequence classification, as in the RDP Classifier (Wang et al., 2007), and k-mer "genomic signatures" of composition (Karlin & Burge, 1995).
• Reference-based tools compared on the benchmark: Kraken2 (taxonomic classification against a sequence database) and Synsor (alignment-free engineered-DNA detection).

License

MIT.