bionode / bionode-vcf Goto Github PK

View Code? Open in Web Editor NEW

5.0 19.0 7.0 47 KB

vcf parser using javascript

License: MIT License

JavaScript 100.00%

bionode tool parser bioinformatics nodejs

bionode-vcf's Introduction

bionode.io

bionode-vcf

a vcf parser in javascript

Install

You need to install the latest Node.JS first, please check nodejs.org or do the following:

# Ubuntu
sudo apt-get install npm
# Mac
brew install node
# Both
npm install -g n
n stable

To use bionode-vcf as a command line tool, you can install it globally with -g.

npm install bionode-vcf -g

Or, if you want to use it as a JavaScript library, you need to install it in your local project folder inside the node_modules directory by doing the same command without -g.

npm i bionode-vcf # 'i' can be used as shortcut to 'install'

Usage

vcf.read

vcf.read takes params: path
The supported filetypes are vcf, zip and gz.

var vcf = require('bionode-vcf');
vcf.read("/path/sample.vcf");
vcf.on('data', function(feature){
    console.log(feature);
})

vcf.on('end', function(){
    console.log('end of file')
})

vcf.on('error', function(err){
    console.error('it\'s not a vcf', err)
})

vcf.readStream

vcf.readStream takes params: stream and extension
The supported extension are vcf, zip and gz.

var vcf = require('bionode-vcf');
var fileStream = s3.getObject({
    Bucket: [BUCKETNAME],
    Key: [FILENAME]
}).createReadStream();  // or stream data from any other source

vcf.read(filestream, 'zip'); // default value is `vcf`
vcf.on('data', function(feature){
    console.log(feature);
})

vcf.on('end', function(){
    console.log('end of file')
})

vcf.on('error', function(err){
    console.error('it\'s not a vcf', err)
})

Documentation

VCF format specifications and more information about the fileds can be found at 1000 genomes webpage and samtools github page

Contributing

We welcome all kinds of contributions at all levels of experience, please read the CONTRIBUTING.md to get started!

bionode-vcf's People

Contributors

Stargazers

Watchers

Forkers

amblina ayangromano marwanzak causecode aneilbaboo a4xrbj1 isabella232

bionode-vcf's Issues

is chai an unused dependency?

chai is listed as a devDependency in the package.json, but the only references to it in the repo are commented out. Can this dependency be removed?

Error: number of columns in the file are less than expected in vcf - there are valid vcf files with just 8 parameters

Please see the attached file format of the Build 37 human genome reference file:

image link

It consists of the following parameters:

#CHROM
POS
ID
REF
ALT
QUAL
FILTER
INFO

This is a valid VCF file, downloaded from Download link

Line 76 in /lib/index.js should have the number 9 change 8 instead:

if (info.length < 9) {

Line 82 in /lib/index.js should also change the number 8 to 7:

var formatIds = info[7].split(':')

to avoid this error thrown. Thank you!

A vcf file with multiple individuals gives duplicate row data

For a vcf file containing data for multiple individuals per line, the the vcf on data event returns correct individual names, but the data for each individual is a duplicate of the first individual's data.
To fix this issue, change line 89 in index.js from this:
var formatParts = info[9].split(':')
to this:
var formatParts = info[9 + j].split(':')

The for loop simply needs to loop to each unique individual in each line.

Add support to parse stream data provided by user

Current implementation takes in path (string) and uses fs module to read file and create stream to further parse it.

REQUEST: A method to parse stream data provided by user should also be available, so user can parse vcf data streamed from any other source.

E.g.: Read data from s3 and parse it directly

const fileStream = s3.getObject({
    Bucket: [BUCKETNAME],
    Key: [FILENAME]
}).createReadStream();

vcf.readStream(fileStream)

Also, it should support compressed files along with plain vcf file.

general question

Hi,

I'm looking for a js module that can read and query compressed and indexed vcf files from a remote location (e.g. S3-buckets). I was wondering if this is the scope of this module. If so, would it be possible to provide some documentation about the possiblities and how to use them?

Alternatively, if my question seems out of scope, do you know about a module that can do such thing?
Thanks
M

possible race condition if setup of event listeners is delayed?

I may just not be used to node/javascript idioms, but I think there is a race condition since there is no way to make sure that the event listeners are set up before events start getting triggered.

For instance, something like the following:

var VCF = require('./lib/index')

var allFeatures = []

var v = VCF.read('test/sample.vcf')

setTimeout(function () {
  v.on('data', (f) => allFeatures.push(f))
  v.on('end', () => console.log(allFeatures))
}, 1000)

will not write out any data because the vcf object emits all of its 'data' events before the 'data' listener is setup. But if the delay parameter (1000) is changed to '0', then the 'data' and 'end' listeners are setup in time.

I don't know if this is a problem in the real-world (if the .on('data'... directly follows the .read call, then maybe there is no way for the race to occur?)

Apologies if I just mis-understand the idiom.

Add support for reading compressed files directly

As the title suggests, add feature to read zip and gzip file directly in addition to reading .vcf files.

Document data callback

The README should document the structure of the feature value provided to the callback in:

vcf.on('data', function (feature) {
   ... // what is feature?
});