Skip to content

Commit

Permalink
Merge pull request #7 from broadinstitute/change_order_and_add_docs_t…
Browse files Browse the repository at this point in the history 
…o_igv_tracks

Changed track order and added docs to igv tracks
  • Loading branch information
@bw2
bw2 authored Sep 12, 2023
2 parents 00fdf95 + 2f6b00d commit 3172a8d
Show file tree
Hide file tree
Showing 2 changed files with 128 additions and 57 deletions.
2 changes: 1 addition & 1 deletion igv.min.js
View file

Large diffs are not rendered by default.

183 changes: 127 additions & 56 deletions index.html
View file
Original file line number Diff line number Diff line change
Expand Up @@ -272,8 +272,8 @@
<br /> -->
<b>Sep 7, 2023</b><br />
- added transcript toggling to simplify SpliceAI transcript output<br />
- added transcript strand to SpliceAI results<br />
- added score visualizations using <a href="https://github.com/bw2/igv.js">custom</a> <a href="https://github.com/igvteam/igv.js">IGV.js</a> tracks</a><br />
- added transcript strand to the results table<br />
- added score visualizations for SpliceAI and Pangolin using <a href="https://github.com/bw2/igv.js">custom IGV.js</a> tracks<br />
<br />
<b>May 22, 2023</b><br />
- changed defaults to 'masked' and max distance to 500bp<br />
Expand Down Expand Up @@ -561,65 +561,105 @@
}

// specify IGV tracks and the data to display in them
const reference = responseSpliceAI.hg == "37" ? "hg19" : "hg38"
const gencodeTrackPath = responseSpliceAI.hg == "37" ?
"gs://tgg-viewer/ref/GRCh37/gencode_v44/knownGene_v44.hg19.sorted.txt.gz" :
"gs://tgg-viewer/ref/GRCh38/gencode_v44/knownGene_v44.hg38.sorted.txt.gz"

const tracks = []

tracks.push({
"name": "Refseq",
"format": "refgene",
"url": responseSpliceAI.hg == "38" ?
"https://hgdownload.soe.ucsc.edu/goldenPath/hg38/database/ncbiRefSeq.txt.gz" :
"https://hgdownload.soe.ucsc.edu/goldenPath/hg19/database/ncbiRefSeq.txt.gz",
"indexed": false,
"infoURL": "https://www.ncbi.nlm.nih.gov/gene/?term=$$"
})

tracks.push({
height: 30,
name: "Variant (?) ",
type: "vcf",
description: `This tracks shows the position of the <b>${chrom}:${pos} ${ref}>${alt}</b> variant.`,
features: [
{
chr: chrom, //see createVCFVariant function in the igv.js repo for the allowed fields
pos: pos,
start: pos - 1,
end: pos,
referenceAllele: ref,
alternateBases: alt,
names: ".", // id in VCF
info: {
variant: `${chrom}-${pos}-${ref}-${alt}`,
}
},
],
})

if (responseSpliceAI) {
tracks.push({
height: 100,
rawOrDelta: "raw",
name: "SpliceAI",
tool: "spliceai",
type: "spliceprediction",
features: responseSpliceAI.allNonZeroScores,
strand: responseSpliceAI.allNonZeroScoresStrand,
name: `SpliceAI (?)`,
description: `<b>Transcript: ${responseSpliceAI.allNonZeroScoresTranscriptId}</b> (${responseSpliceAI.allNonZeroScoresStrand == '-' ? 'minus' : 'plus'} strand)<br />
SpliceAI predictions of splice acceptor & donor sites within the +/- ${responseSpliceAI.distance}bp window<br />
around <b>${chrom}:${pos} ${ref}>${alt}</b>. This track shows markers that represent the scores for the<br/>
<span style="color:#0000B4"><b>reference sequence</b></span> (without the variant) and the<br />
<span style="color:#05d0d2"><b>alternate sequence</b></span> (with the variant). <br />
A marker is shown at all positions where a score is ≥ 0.01. <br/>
SpliceAI outputs separate predictions for whether a given position is a splice acceptor and/or a splice donor, so<br/>
<b>A</b>'s represent acceptor probability scores, and <br/>
<b>D</b>'s represent donor probability scores.<br/>
The letters are shown up-side-down when the alternate sequence score is lower than the reference sequence score.<br/>
Numerical labels show the scores predicted from the reference sequence.
`,
height: 100,
rawOrDelta: "raw",
tool: "SpliceAI",
type: "spliceprediction",
features: responseSpliceAI.allNonZeroScores,
strand: responseSpliceAI.allNonZeroScoresStrand,
})
tracks.push({
height: 200,
rawOrDelta: "delta",
name: "SpliceAI",
tool: "spliceai",
type: "spliceprediction",
features: responseSpliceAI.allNonZeroScores,
strand: responseSpliceAI.allNonZeroScoresStrand,
name: `SpliceAI (?)`,
description: `<b>Transcript: ${responseSpliceAI.allNonZeroScoresTranscriptId}</b> (${responseSpliceAI.allNonZeroScoresStrand == '-' ? 'minus' : 'plus'} strand)<br />
SpliceAI delta score predictions for gain and loss of splice acceptors and donors within a <br />
+/- ${responseSpliceAI.distance}bp window around <b>${chrom}:${pos} ${ref}>${alt}</b>.<br />
This track shows markers at all positions where the |delta score| is ≥ 0.01. <br/>
<b>Bar colors:</b> <b style="color:#FF0000">red</b> is for delta scores ≥ 0.8,
<b style="color:#FFCB1F">yellow</b> is for delta scores ≥ 0.5,
<b style="color:#1fb839">green</b> is for delta scores ≥ 0.2<br />
<b>A</b>'s represent acceptor probability scores<br/>
<b>D</b>'s represent donor probability scores<br/>
`,
height: 200,
rawOrDelta: "delta",
tool: "SpliceAI",
type: "spliceprediction",
features: responseSpliceAI.allNonZeroScores,
strand: responseSpliceAI.allNonZeroScoresStrand,
})
}

if (responsePangolin) {
tracks.push({
name: `Pangolin (?)`,
description: `<b>Gene: ${responsePangolin.allNonZeroScoresTranscriptId}</b> (${responsePangolin.allNonZeroScoresStrand == '-' ? 'minus' : 'plus'} strand)<br />
Pangolin delta score predictions for gain and loss of splice sites at positions<br />
within a +/- ${responsePangolin.distance}bp window around <b>${chrom}:${pos} ${ref}>${alt}</b>.<br />
This track shows markers at all positions where the delta score is ≥ 0.1. <br/>
<b>Bar colors:</b> <b style="color:#FF0000">red</b> is for delta scores ≥ 0.8,
<b style="color:#FFCB1F">yellow</b> is for delta scores ≥ 0.5,
<b style="color:#1fb839">green</b> is for delta scores ≥ 0.2<br />
`,
height: 200,
rawOrDelta: "delta",
name: "Pangolin",
tool: "pangolin",
tool: "Pangolin",
type: "spliceprediction",
features: responsePangolin.allNonZeroScores,
strand: responsePangolin.allNonZeroScoresStrand,
})
}

tracks.push({
height: 30,
name: "Variant",
type: "vcf",
features: [
{
chr: chrom, //see createVCFVariant function in the igv.js repo for the allowed fields
pos: pos,
start: pos - 1,
end: pos,
referenceAllele: ref,
alternateBases: alt,
names: ".", // id in VCF
info: {
variant: `${chrom}-${pos}-${ref}-${alt}`,
}
},
],
})
const gencodeTrackPath = responseSpliceAI.hg == "37" ?
"gs://tgg-viewer/ref/GRCh37/gencode_v44/knownGene_v44.hg19.sorted.txt.gz" :
"gs://tgg-viewer/ref/GRCh38/gencode_v44/knownGene_v44.hg38.sorted.txt.gz"

tracks.push({
name: "Gencode v44",
Expand All @@ -640,9 +680,13 @@
for (const splicePredictionType of ["loss", "gain"]) {
tracks.push(
{
height: 100,
name: `SpliceAI: splice ${splicePredictionType} score >= ${minScore}`,
name: `SpliceAI: A or D ${splicePredictionType} ≥ ${minScore} (?)`,
description: `This track visualizes Illumina's precomputed SpliceAI score tables for SNVs and small INDELs.<br/>
Each genomic location where an SNV or INDEL variant would have a SpliceAI score ≥ ${minScore} is shown as the origin of an arrow.<br />
The arrow points to the location where that variant would most likely cause acceptor or donor ${splicePredictionType}.
Clicking on the arrow will display additional information.`,
type: "spliceJunctions",
height: 100,
url: `gs://tgg-viewer/ref/GRCh38/spliceai/spliceai_scores.raw.snps_and_indels.hg38.filtered.sorted.score_${minScore}.splice_${splicePredictionType}.bed.gz`,
indexURL: `gs://tgg-viewer/ref/GRCh38/spliceai/spliceai_scores.raw.snps_and_indels.hg38.filtered.sorted.score_${minScore}.splice_${splicePredictionType}.bed.gz.tbi`,
}
Expand All @@ -651,16 +695,19 @@
}


for (const filenamePrefix of [
"GTEX_muscle.803_samples",
"GTEX_blood.755_samples",
"GTEX_fibs.504_samples",
"GTEX_lymphocytes.174_samples",
"GTEX_brain_cortex.255_samples",
//"GTEX_frontal_cortex.209_samples",
for (const [filenamePrefix, tissue, sampleCount] of [
["GTEX_muscle.803_samples", "muscle", 803],
["GTEX_blood.755_samples", "blood", 755],
["GTEX_fibs.504_samples", "fibroblasts", 504],
["GTEX_lymphocytes.174_samples", "lymphocytes", 174],
["GTEX_brain_cortex.255_samples", "brain cortex", 255],
//["GTEX_frontal_cortex.209_samples", "frontal cortex", 803],
]) {
tracks.push({
name: filenamePrefix.replaceAll(".", ": ").replaceAll("_", " ").replace("fibs", "fibroblasts"),
name: `GTEx ${tissue}: ${sampleCount} samples (?)`,
description: `This track shows the combined splice junctions from all ${sampleCount} ${tissue} samples available in GTEx v8.
Splice junctions are labeled with the total number of RNA-seq reads that supported the junction, summed across the ${sampleCount} samples.
The coverage track shows the total number of RNA-seq reads that overlap each position, summed across all samples and divided by ${sampleCount}.`,
type: "merged",
height: 100,
tracks: [{
Expand All @@ -679,10 +726,34 @@
})
}

let locusMargin = maxPos - minPos < 200 ? 15 : 50

// igv.js reference genome specs are copied from https://igv.org/genomes/genomes.json
// These must be specified explicitly rather than just setting the reference to "hg19" or "hg38" so that
// the RefSeq gene track can be the first track (as described in https://github.com/igvteam/igv.js/issues/1518)
let reference = responseSpliceAI.hg == "37" ? {
"id": "hg19",
"name": "Human (GRCh37/hg19)",
"fastaURL": "https://igv-genepattern-org.s3.amazonaws.com/genomes/seq/hg19/hg19.fasta",
"indexURL": "https://igv-genepattern-org.s3.amazonaws.com/genomes/seq/hg19/hg19.fasta.fai",
"cytobandURL": "https://igv-genepattern-org.s3.amazonaws.com/genomes/seq/hg19/cytoBand.txt",
"aliasURL": "https://s3.amazonaws.com/igv.org.genomes/hg19/hg19_alias.tab",
"chromosomeOrder": "chr1, chr2, chr3, chr4, chr5, chr6, chr7, chr8, chr9, chr10, chr11, chr12, chr13, chr14, chr15, chr16, chr17, chr18, chr19, chr20, chr21, chr22, chrX, chrY",
"tracks": tracks,
} : {
"id": "hg38",
"name": "Human (GRCh38/hg38)",
"fastaURL": "https://igv-genepattern-org.s3.amazonaws.com/genomes/seq/hg38/hg38.fa",
"indexURL": "https://igv-genepattern-org.s3.amazonaws.com/genomes/seq/hg38/hg38.fa.fai",
"cytobandURL": "https://s3.amazonaws.com/igv.org.genomes/hg38/annotations/cytoBandIdeo.txt.gz",
"aliasURL": "https://s3.amazonaws.com/igv.org.genomes/hg38/hg38_alias.tab",
"chromosomeOrder": "chr1, chr2, chr3, chr4, chr5, chr6, chr7, chr8, chr9, chr10, chr11, chr12, chr13, chr14, chr15, chr16, chr17, chr18, chr19, chr20, chr21, chr22, chrX, chrY",
"tracks": tracks,
}

return {
reference: reference,
locus: `${chrom}:${minPos - 50}-${maxPos + 50}`,
tracks: tracks,
locus: `${chrom}:${minPos - locusMargin}-${maxPos + locusMargin}`,
}
}

Expand Down Expand Up @@ -736,8 +807,6 @@
//sort by (gene name, is canonical transcript, is non-coding transcriipt)
parsed_scores_or_errors = _.chain(parsed_scores_or_errors).sortBy('sort_key').value()

console.log("Parsed scores:", parsed_scores_or_errors)

return parsed_scores_or_errors.map(function(scores_or_error, index) {

let scores = scores_or_error
Expand All @@ -760,6 +829,8 @@
const biotype_link = `<div class="smallLink"><a href="https://www.gencodegenes.org/pages/biotypes.html" target="_blank">${transcript_type.replace(/_/g, " ")}</a></div>`
const refseq_transcript_id_link = refseq_transcript_ids ? (" /" + ` <a href="https://www.ncbi.nlm.nih.gov/search/all/?term=${refseq_transcript_ids[0]}" target="_blank">${refseq_transcript_ids[0]}</a>`) : ''
gene_and_transcript_id = ` <div class="smallLink">(${gene_id_link}${transcript_id_link}${refseq_transcript_id_link})</div><br /><br class='only-large-screen' />${biotype_link}${canonical_transcript_text} <div class="smallLink"> (${strand} strand)</div>`
} else {
gene_and_transcript_id = ` <div class="smallLink"> (${strand} strand)</div>`
}
//round to 2 decimal places
if (scores.is_pangolin) {
Expand Down

0 comments on commit 3172a8d

Please sign in to comment.