hic-straw

May 14, 2026 · View on GitHub

CI

Command line and web utilities for reading .hic contact matrix files, with support for creating live contact and distance maps from 3D chromatin structure data for Spacewalk

Installation

Requires Node 18+ (https://nodejs.org)

npm install hic-straw

API

getContactRecords

Return a collection of binned contact counts.

Arguments

  • normalization - string indicating normalization scheme
  • region 1 {chr, start, end} - genomic region in base pair or fragment units. Interval convention is zero based 1/2 open
  • region 2 {chr, start, end}
  • units -- "BP" for base pairs. Currently this is the only unit supported
  • binSize -- size of each bin in base pair or fragment units. Bins are square

Examples

Development

Run the examples dashboard with hot reload:

npm run dev

Then open http://localhost:5173 in your browser. The dashboard links to each example.

Browser usage

ES module — see examples/straw.html

import Straw from 'hic-straw'

const straw = new Straw({
    url: "https://s3.amazonaws.com/igv.broadinstitute.org/data/hic/intra_nofrag_30.hic"
})

straw.getContactRecords(
    "KR",
    { chr: "8", start: 50000000, end: 60000000 },
    { chr: "8", start: 50000000, end: 60000000 },
    "BP",
    1000000
)
    .then(function (contactRecords) { ... })

Node

local file

To use hic-straw with a local file, use the NodeLocalFile class:

import Straw from 'hic-straw'
import NodeLocalFile from 'hic-straw/node'

const nodeLocalFile = new NodeLocalFile({ path: "test/data/test_chr22.hic" })
const straw = new Straw({ file: nodeLocalFile })

remote file

Node 18+ includes native fetch. For remote files:

import Straw from 'hic-straw'

const straw = new Straw({ url: "https://foo.bar/test.hic" })

Command line

Note: "straw" is installed in node_modules/.bin/straw. This should be added to the path automatically upon installing hic-straw, however if you get the error straw: command not found try running straw explicitly as

node_modules/.bin/straw...

Extract file metadata (genome identifier, sequences, resolutions)

straw --meta test/data/test_chr22.hic

Extract normalization options.

straw --norms test/data/test_chr22.hic

Extract contact records from a local hic file


straw KR test/data/test_chr22.hic 22:40,000,000-50,000,000 22:40,000,000-50,000,000 BP 100,000

Extract contact records from a remote hic file

straw KR https://s3.amazonaws.com/igv.broadinstitute.org/data/hic/intra_nofrag_30.hic 8:48,700,000-48,900,000 8:48700000-48900000 BP 10,000

LiveContactMap — Synthetic Contact Maps from 3D Structure Data

LiveContactMap is an adapter that accepts 3D chromosome vertex data and produces contact maps fully compatible with the hic-straw / Juicebox.js pipeline. It implements the same interface as HicFile, so downstream consumers (Straw, Juicebox) cannot distinguish it from a real .hic file.

This is designed for Spacewalk, which visualizes 3D chromatin tracing data and needs to display live contact and distance maps alongside the 3D structure.

How it works

  1. Parse 3D vertex data from a Spacewalk Text (SWT) file or provide traces directly
  2. Compute a pairwise Euclidean distance matrix, averaged across all traces in an ensemble
  3. Derive contact records by applying a distance threshold — pairs closer than the threshold are "in contact". In frequency mode, the count reflects the fraction of traces where the pair is in contact (0.0 to 1.0)
  4. Serve contact records through the standard HicFile interface

Supported input formats

  • SWT (.swt) — text ball-and-stick format, see below.
  • SW / SWB (.sw, .swb) — binary HDF5 ball-and-stick format used by Spacewalk. V1 supports SINGLE_POINT files with a single ensemble group and a single genomic region. Pointcloud (MULTI_POINT) files are not yet supported.

Input: SWT file format

Spacewalk Text files (.swt) describe ball-and-stick models of chromatin fiber. Each file contains multiple traces (independent 3D conformations) of the same genomic region.

##format=sw1 name=IMR90 genome=hg38
chromosome	start	end	x	y	z
trace 0
chr21 18000000 18030000 117803 58446 1733
chr21 18030000 18060000 117726 58747 1680
chr21 18060000 18090000 117747 58607 1872
...
trace 1
chr21 18000000 18030000 ...
...

See resources/spacewalk-swt-text-file-format.md for the full format specification.

API

Construction

There are three ways to create a LiveContactMap:

From SWT text (simplest — parses the file for you):

import LiveContactMap from 'hic-straw/src/liveContactMap.js'

const swtText = fs.readFileSync('data/ball-and-stick.swt', 'utf-8')
const lcm = new LiveContactMap({
    swtText: swtText,
    distanceThreshold: 500,   // 3D distance cutoff; omit to derive from the data
    contactMode: 'frequency'  // 'frequency' (0-1) or 'contact' (binary 0/1)
})
await lcm.init()

From a binary SW / SWB (HDF5) file — browser File, remote URL, or Node path:

// Browser — File picker or drag-drop
const lcm = new LiveContactMap({ swFile: fileObject, distanceThreshold: 500 })

// Remote URL (uses HTTP range requests — no full download required)
const lcm = new LiveContactMap({ swUrl: 'https://host/data.sw', distanceThreshold: 500 })

// Node
const lcm = new LiveContactMap({ swPath: 'data/ball-and-stick.sw', distanceThreshold: 500 })

await lcm.init()

From raw trace data (when Spacewalk already has parsed vertex arrays):

const lcm = new LiveContactMap({
    traces: ensembleManager.getTraces(),   // Array<Array<{x, y, z}>>
    genomeId: 'hg38',
    chr: 'chr21',
    genomicStart: 18000000,
    genomicEnd: 19950000,
    binSize: 30000,
    distanceThreshold: 500
})
await lcm.init()

From pre-parsed SWT data (if you've already called parseSWT()):

import { parseSWT } from 'hic-straw/src/swtParser.js'

const parsed = parseSWT(swtText)
const lcm = new LiveContactMap({
    parsedData: parsed,
    distanceThreshold: 500
})
await lcm.init()

Querying contact records

After initialization, use the standard hic-straw interface:

const records = await lcm.getContactRecords(
    'NONE',
    { chr: 'chr21', start: 18000000, end: 19950000 },
    { chr: 'chr21', start: 18000000, end: 19950000 },
    'BP',
    30000
)

for (const rec of records) {
    console.log(`bin ${rec.bin1} x ${rec.bin2}: ${rec.counts}`)
}

Using with Straw (for Juicebox compatibility)

To plug a LiveContactMap into the Straw/Juicebox pipeline:

import Straw from 'hic-straw'

const straw = new Straw({ liveContactMap: lcm })

// Now use straw exactly like a normal .hic file:
const meta = await straw.getMetaData()
const records = await straw.getContactRecords('NONE', region1, region2, 'BP', 30000)

Juicebox's HiCDataset can wrap this Straw instance and all controls (resolution selector, normalization widget, color scale) work natively.

Dynamic threshold adjustment

Changing the distance threshold re-derives contacts without recomputing the expensive distance matrix:

lcm.setDistanceThreshold(300)    // fewer contacts
lcm.setDistanceThreshold(800)    // more contacts

Accessing the distance matrix

For distance map visualization:

const { distances, maxDistance, traceLength } = lcm.getDistanceMatrix()
// distances: Float32Array (N x N, row-major, symmetric)
// maxDistance: largest distance in the matrix
// traceLength: N (number of bins)

Node.js example

import fs from 'fs'
import Straw from 'hic-straw'
import LiveContactMap from 'hic-straw/src/liveContactMap.js'

const swtText = fs.readFileSync('resources/ball-and-stick.swt', 'utf-8')

const lcm = new LiveContactMap({
    swtText,
    distanceThreshold: 500,
    contactMode: 'frequency'
})
await lcm.init()

const meta = await lcm.getMetaData()
console.log(`Genome: ${meta.genome}`)
console.log(`Chromosomes: ${meta.chromosomes.map(c => c.name).join(', ')}`)
console.log(`Resolution: ${meta.resolutions[0]} bp`)

const records = await lcm.getContactRecords(
    'NONE',
    { chr: 'chr21', start: 18000000, end: 19950000 },
    { chr: 'chr21', start: 18000000, end: 19950000 },
    'BP',
    30000
)
console.log(`Contact records: ${records.length}`)

Visual test page

A browser-based test page is provided at examples/live-contact-map.html. It renders both a contact map and a distance map side by side from any .swt file, with interactive controls for distance threshold and contact mode.

Run npm run dev and open the examples dashboard, then click "LiveContactMap" — or navigate directly to http://localhost:5173/examples/live-contact-map.html. Load an SWT file using the file picker.