cyanea-seq

v0.3.8 Core

The sequence toolkit at the heart of Cyanea.

Core layer Apache-2.0 10 functions Interactive playground

Biological sequence types and operations — DNA, RNA, protein. Transcription, translation, k-mers, reverse complement, GC content, RNA folding, and more.

Playground

Loading playground…

Overview

cyanea-seq is the most-used crate in the ecosystem. It provides strongly-typed biological sequence representations (DNA, RNA, protein) with efficient 2-bit encoding, plus a comprehensive toolkit of sequence operations: transcription, translation, reverse complement, GC content, k-mer generation, MinHash similarity, RNA secondary structure prediction, and contig assembly statistics.

Every operation that manipulates a raw biological sequence lives here, making it the natural dependency for any crate that reads, writes, aligns, or analyzes sequences.

Key Concepts

2-Bit Encoding

DNA bases require only 2 bits each (A=00, C=01, G=10, T=11). cyanea-seq stores sequences in packed form, reducing memory by 4x compared to one-byte-per-base representations. All operations work directly on the packed representation — no unpacking needed.

The Central Dogma

transcribe converts a DNA template into mRNA (T → U). translate reads the mRNA three bases at a time, mapping each codon to its amino acid via the standard genetic code. Together they implement the central dogma of molecular biology in two function calls.

K-mer Analysis

K-mers are all subsequences of length k in a sequence. They are the foundation of genome assembly, metagenomic classification, and sequence comparison. kmer_count produces a frequency table; minhash_compare uses locality-sensitive hashing to estimate Jaccard similarity between two sets of k-mers in sublinear time.

RNA Secondary Structure

rna_fold_nussinov implements the Nussinov algorithm to predict the maximum number of base pairs in an RNA sequence. The result is a dot-bracket string where ( and ) indicate paired bases and . indicates unpaired positions.

Code Examples

Rust

use cyanea_seq::{DnaSeq, transcribe, translate, gc_content};

let seq = DnaSeq::from("ATGGCTAGCAAAGAC");
let mrna = transcribe(&seq);
let protein = translate(&seq);
let gc = gc_content(&seq);

Python

import cyanea

mrna = cyanea.transcribe("ATGGCTAGCAAAGAC")
protein = cyanea.translate("ATGGCTAGCAAAGAC")
gc = cyanea.gc_content("ATGGCTAGCAAAGAC")
fold = cyanea.rna_fold_nussinov("GGGAAACCC")

JavaScript (WASM)

import { transcribe, translate, gc_content_json, rna_fold_nussinov } from '/wasm/cyanea_wasm.js';

const mrna = transcribe("ATGGCTAGCAAAGAC");
const protein = translate("ATGGCTAGCAAAGAC");
const gc = JSON.parse(gc_content_json("ATGGCTAGCAAAGAC"));
const fold = JSON.parse(rna_fold_nussinov("GGGAAACCC"));

Use Cases

Gene annotation — Translate ORFs and compute reading frame statistics.
Quality control — Validate sequences against expected alphabets before analysis.
Metagenomic binning — Compare k-mer profiles between contigs via MinHash.
RNA structure — Predict simple secondary structures for short non-coding RNAs.

API Surface

transcribe (seq: &str) -> String Convert DNA to mRNA (T → U)

translate (seq: &str) -> String Translate DNA/RNA to amino acid sequence

reverse_complement (seq: &str) -> String Reverse complement of a DNA sequence

gc_content_json (seq: &str) -> f64 Calculate GC content as a fraction

validate (seq: &str, alphabet: &str) -> bool Check sequence validity against an alphabet

rna_fold_nussinov (seq: &str) -> String Predict RNA secondary structure (Nussinov algorithm)

protein_props (seq: &str) -> JSON Compute protein physicochemical properties

codon_usage (seq: &str) -> JSON Compute codon usage table from a coding sequence

minhash_compare (a: &str, b: &str, k, n) -> f64 Estimate Jaccard similarity via MinHash sketches

assembly_stats_json (contigs: JSON) -> JSON Compute N50, L50, and assembly statistics

Depends on

cyanea-core

Depended on by

cyanea-io cyanea-align cyanea-omics cyanea-stats cyanea-wasm