Crate oricle 

Bacterial oriC (replication origin) prediction — DB-free, pure computation.

Method (Ori-Finder style): the cumulative GC-skew (Z-curve) of a bacterial chromosome falls along the replichore running from the terminus to the origin and rises along the other, so its global minimum sits at oriC. The minimum is refined with clusters of the DnaA-box 9-mer (TTATCCACA consensus, plus reverse complement) and, when supplied, proximity to the dnaA gene call.

No database, no network, no alignment: detect takes bytes and returns coordinates.

use oricle::{detect, GeneHint};

let hits = detect(&genome, &[]);          // gene hints are optional
for o in &hits {
    println!("oriC {}..{} (score {:.2})", o.start, o.end, o.score);
}

Scope and limits

• Intended for complete or near-complete replicons. The skew curve is meaningless on short draft contigs, so sequences below Options::min_len are rejected outright.
• Assumes a circular replicon by default. For a linear chromosome (Streptomyces, Borreliella) set Options::topology to Topology::Linear, which disables the antipodal terminus check.
• Some lineages genuinely have a flat or noisy skew (cyanobacteria, Deinococcus). Those are reported as no call rather than guessed at; supply a dnaA GeneHint to enable the gene-anchored fallback.
• Plasmids do not carry a DnaA-dependent oriC. The signal gates are tuned to leave them uncalled.

Structs

GeneHint

A predicted gene position, so the caller can supply the dnaA location for proximity refinement. Optional — pass an empty slice to skip.

Options

Tuning knobs for detect_with. Start from Options::default.

OriC

A predicted replication origin.

Enums

Confidence

Coarse, action-oriented confidence class for a call, for quick triage and filtering. Derived from OriC::score against thresholds calibrated on the full held-out set so the class tracks real ≤5 kb accuracy: Pass calls are right the large majority of the time, Weak calls are wrong more often than not, and Review sits in between. The OriC::notes say why.

Flag

A qualitative note attached to a call: either a warning that should lower a reader’s trust, or a contextual fact worth surfacing. Zero or more apply to any call; they explain the OriC::conf class. Each has a stable code for the CLI notes column.

Method

How a call was arrived at.

Topology

Replicon topology. Bacterial chromosomes are usually circular.

Functions

detect

Detect oriC in one replicon with default Options (circular topology).

detect_with

Detect oriC in one replicon with explicit Options.