Nucleotide sequence tracks for HiGlass

Display nucleotide sequences in HiGlass!

Note: This is the source code for the sequence track only! You might want to check out the following repositories as well:

• HiGlass viewer: https://github.com/higlass/higlass
• HiGlass server: https://github.com/higlass/higlass-server
• HiGlass docker: https://github.com/higlass/higlass-docker

Installation

npm install higlass-sequence

Usage

The live script can be found at:

• https://unpkg.com/@reservoirgenomics/[email protected]/dist/higlass-sequence.js

Client

1. Make sure you load this track prior to hglib.js. For example:

<script src="/higlass-sequence.js"></script>
<script src="hglib.js"></script>
<script>
  ...
</script>

2. higlass-sequence can load sequence data in two ways:

• Convert a FASTA file into the multivec format. Use this format if you need aggregation on lower zoom levels. See the colorAggregationMode option below. The script scripts/fasta_to_hdf5.py will convert your FASTA file into the hdf5 format which then can be converted to the multivec format (see the HiGlass docs for more information). After ingesting the multivec file into a HiGlass server, you can configure the track in your view config in the usual way, using a server and tilesetUid:

{
  "server": "http://localhost:8001/api/v1",
  "tilesetUid": "ScrlBGMbR_WJ0fzKXxKFzA",
  "uid": "seq_multivec_example",
  "type": "horizontal-sequence",
  "options": {
    ...
  },
  "width": 568,
  "height": 25
},

• The sequence track can also load the data directly into the client using indexed FASTA files. No running HiGlass server is required in this case. The only additional data required (besides the FASTA file) is a fai index file and a chrom sizes file. The track then can be configured in the following way:

{
 "uid": "seq_fasta_example",
 "type": "horizontal-sequence",
 "data": {
   "type": "fasta",
   "fastaUrl": "https://aveit.s3.amazonaws.com/higlass/data/sequence/hg38.fa",
   "faiUrl": "https://aveit.s3.amazonaws.com/higlass/data/sequence/hg38.fa.fai",
   "chromSizesUrl": "https://aveit.s3.amazonaws.com/higlass/data/sequence/hg38.mod.chrom.sizes"
 },
 "options": {
   ...
 },
 "width": 568,
 "height": 25
},

Using indexed FASTA files is easier and faster than multivec files. If no aggregation is needed, this is the preferred way to load data in higlass-sequence.

For an example, see src/index.html.

3. The following options are available:

{
   "colorAggregationMode": "none", // ['none', 'max', 'weighted']. Ignored when using indexed FASTA files. Determines how pixels are colored on lower zoom levels. 'max' colors the pixels according to the nucleotide with the highest occurence. 'weighted' uses a weighted interpolation of colors.
   "extendedPreloading": false, // More tiles than necessary are loaded. Leads to a smoother horizontal scrolling especially for multivec files.
   "labelPosition": "topLeft",
   "labelColor": "black",
   "labelTextOpacity": 0.4,
   "notificationText": "Zoom in to see nucleotides...", // Notification text if colorAggregationMode == "none" and user zooms out too far.
   "trackBorderWidth": 0,
   "trackBorderColor": "white",
   "backgroundColor": "white",
   "barBorder": true,
   "barBorderColor": "white",
   "fontSize": 16, // size of the letters
   "fontFamily": "Arial",
   "fontColor": "white",
   "colorScale": [
     "#007FFF", // color of A
     "#e8e500", // color of T
     "#008000", // color of G
     "#FF0038", // color of C
     "#800080", // color of N
     "#DCDCDC", // color of everything else
   ],
 }

ECMAScript Modules (ESM)

We also build out ES modules for usage by applications who may need to import or use higlass-sequence as a component.

Whenever there is a statement such as the following, assuming higlass-sequence is in your node_modules folder:

import { SequenceTrack } from 'higlass-sequence';

Then SequenceTrack would automatically be imported from the ./es directory (set via package.json's "module" value).

Support

For questions, please either open an issue or ask on the HiGlass Slack channel at http://bit.ly/higlass-slack

Development

Testing

To run the test suite:

npm run test-watch

Installation

$ git clone https://github.com/higlass/higlass-sequence.git
$ cd higlass-sequence
$ npm install

If you have a local copy of higlass, you can then run this command in the higlass-sequence directory:

npm link higlass

Commands

• Developmental server: npm start
• Production build: npm run build