@sjcrh/proteinpaint-shared 2.78.0-0

package/src/bulk.snv.js

@@ -0,0 +1,234 @@
+///////////////////////////////
+//
+//  shared between client and server
+//
+///////////////////////////////
+
+import * as common from './common.js'
+import * as bulk from './bulk.js'
+
+export function parseheader(line, flag) {
+	const header = line.toLowerCase().split('\t')
+	if (header.length <= 1) return 'invalid file header for snv/indel'
+	const htry = (...args) => {
+		for (const s of args) {
+			const i = header.indexOf(s)
+			if (i != -1) return i
+		}
+		return -1
+	}
+	let i = htry('annovar_gene', 'annovar_sj_gene', 'gene', 'genename', 'gene_symbol', 'hugo_symbol')
+	if (i == -1) return 'gene missing from header'
+	header[i] = 'gene'
+	i = htry('annovar_aachange', 'amino_acid_change', 'annovar_sj_aachange', 'aachange', 'protein_change', 'variant')
+	if (i == -1) return 'amino_acid_change missing from header'
+	header[i] = 'mname'
+	i = htry('annovar_class', 'class', 'mclass', 'variant_class', 'variant_classification', 'annovar_sj_class')
+	if (i == -1) return 'variant_class missing from header'
+	header[i] = 'class'
+	i = htry('chromosome', 'chr')
+	if (i == -1) return 'chromosome missing from header'
+	header[i] = 'chr'
+	i = htry('wu_hg19_pos', 'start', 'start_position', 'chr_position', 'position')
+	if (i == -1) return 'start missing from header'
+	header[i] = 'pos'
+	i = htry(
+		'annovar_isoform',
+		'mrna_accession',
+		'mrna accession',
+		'refseq_mrna_id',
+		'annovar_sj_filter_isoform',
+		'refseq',
+		'isoform'
+	)
+	if (i == -1) return 'isoform missing from header'
+	header[i] = 'isoform'
+
+	// optional
+	i = htry('sample', 'sample_name', 'tumor_sample_barcode')
+	if (i != -1) header[i] = 'sample'
+	i = htry('patient', 'donor', 'target_case_id')
+	if (i != -1) header[i] = 'patient'
+	i = htry('quantitative_measurements')
+	if (i != -1) header[i] = 'qmset'
+	// dna maf tumor
+	i = htry('mutant_reads_in_case', 'mutant_in_tumor', 'tumor_readcount_alt')
+	if (i != -1) header[i] = 'maf_tumor_v1'
+	i = htry('total_reads_in_case', 'total_in_tumor', 'tumor_readcount_total')
+	if (i != -1) header[i] = 'maf_tumor_v2'
+	// dna maf normal
+	i = htry('mutant_reads_in_control', 'mutant_in_normal', 'normal_readcount_alt')
+	if (i != -1) header[i] = 'maf_normal_v1'
+	i = htry('total_reads_in_control', 'total_in_normal', 'normal_readcount_total')
+	if (i != -1) header[i] = 'maf_normal_v2'
+	// rna maf
+	// cdna
+	i = htry('cdna_change')
+	if (i != -1) header[i] = 'cdna_change'
+	i = htry('sampletype', 'sample type', 'sample_type')
+	if (i != -1) header[i] = 'sampletype'
+	i = htry('origin')
+	if (i != -1) header[i] = 'origin'
+	i = htry('cancer', 'disease', 'diagnosis')
+	if (i != -1) header[i] = 'disease'
+	flag.snv.header = header
+	flag.snv.loaded = true
+	return false
+}
+
+export function parseline(linei, line, flag) {
+	if (line == '' || line[0] == '#') return
+	const lst = line.split('\t')
+	const m = {}
+	for (let j = 0; j < flag.snv.header.length; j++) {
+		if (lst[j] == undefined) break
+		m[flag.snv.header[j]] = lst[j]
+	}
+	if (!m.gene) {
+		flag.snv.badlines.push([linei, 'missing gene', lst])
+		return
+	}
+	if (m.gene.toUpperCase() == 'UNKNOWN') {
+		flag.snv.badlines.push([linei, 'gene name is UNKNOWN', lst])
+		return
+	}
+	if (!m.isoform) {
+		flag.snv.badlines.push([linei, 'missing isoform', lst])
+		return
+	}
+	if (!m.mname) {
+		m.mname = m.cdna_change
+		if (!m.mname) {
+			flag.snv.badlines.push([linei, 'missing amino acid change', lst])
+			return
+		}
+	} else {
+		if (m.mname.indexOf('p.') == 0) {
+			m.mname = m.mname.replace(/^p\./, '')
+		}
+	}
+	if (!m.class) {
+		flag.snv.badlines.push([linei, 'missing mutation class', lst])
+		return
+	}
+	let _c = flag.mclasslabel2key[m.class.toUpperCase()]
+	if (_c) {
+		m.class = _c
+	} else {
+		_c = common.mclasstester(m.class)
+		if (_c) {
+			m.class = _c
+		} else {
+			flag.snv.badlines.push([linei, 'wrong mutation class: ' + m.class, lst])
+			return
+		}
+	}
+	if (bulk.parsesample(m, flag, linei, lst, flag.snv.badlines)) {
+		return
+	}
+	if (!m.chr) {
+		flag.snv.badlines.push([linei, 'missing chromosome', lst])
+		return
+	}
+	if (m.chr.toLowerCase().indexOf('chr') != 0) {
+		m.chr = 'chr' + m.chr
+	}
+	if (!m.pos) {
+		flag.snv.badlines.push([linei, 'missing chromosome position', lst])
+		return
+	}
+	const v = Number.parseInt(m.pos)
+	if (Number.isNaN(v)) {
+		flag.snv.badlines.push([linei, 'invalid chromosome position', lst])
+		return
+	}
+	m.pos = v - 1
+
+	if (m.maf_tumor_v2 != undefined && m.maf_tumor_v1 != undefined) {
+		if (m.maf_tumor_v2 == '') {
+			// no value, do not parse
+		} else {
+			let v1 = Number.parseInt(m.maf_tumor_v1),
+				v2 = Number.parseInt(m.maf_tumor_v2)
+			if (Number.isNaN(v1) || Number.isNaN(v2)) {
+				flag.snv.badlines.push([linei, 'invalid maf_tumor mutant and/or total read count', lst])
+				return
+			}
+			m.maf_tumor = { f: v1 / v2, v1: v1, v2: v2 }
+		}
+		delete m.maf_tumor_v1
+		delete m.maf_tumor_v2
+	}
+
+	if (m.maf_normal_v1 != undefined && m.maf_normal_v2 != undefined) {
+		if (m.maf_normal_v2 == '') {
+			// no value
+		} else {
+			let v1 = Number.parseInt(m.maf_normal_v1),
+				v2 = Number.parseInt(m.maf_normal_v2)
+			if (Number.isNaN(v1) || Number.isNaN(v2)) {
+				flag.snv.badlines.push([linei, 'invalid maf_normal mutant and/or total read count', lst])
+				return
+			}
+			m.maf_normal = { f: v1 / v2, v1: v1, v2: v2 }
+		}
+		delete m.maf_normal_v1
+		delete m.maf_normal_v2
+	}
+
+	/*
+	if(m.qmset) {
+		try{
+			var v=JSON.parse(m.qmset)
+		} catch(e){
+			flag.snv.badlines.push([linei,'invalid JSON for quantitative_measurements',lst])
+			v=null
+		} finally {
+			if(v) {
+				if(typeof(v)!='object') {
+					flag.snv.badlines.push([linei,'value of quantitative_measurements must be an object',lst])
+					delete m.qmset
+				} else {
+					for(var n in v) {
+						if(!Array.isArray(v[n])) {
+							flag.snv.badlines.push([linei,'quantitative_measurements: "'+n+'" value must be an array',lst])
+							delete v[n]
+						} else {
+							var tmp=[]
+							v[n].forEach(function(v2){
+								if(typeof(v2)=='number') {
+									tmp.push({v:v2})
+								} else if(v2.v && typeof(v2.v)=='number') {
+									tmp.push(v2)
+								}
+							})
+							if(tmp.length) {
+								v[n]=tmp
+							} else {
+								flag.snv.badlines.push([linei,'quantitative_measurements: no valid value for "'+n+'"',lst])
+								delete v[n]
+							}
+						}
+					}
+					m.qmset=v
+				}
+			} else {
+				delete m.qmset
+			}
+		}
+	}
+	*/
+	flag.good++
+	// FIXME hard-coded M and S
+	if (m.class == 'M') {
+		flag.snv.missense++
+	} else if (m.class == 'S') {
+		flag.snv.silent++
+	}
+	const n = flag.geneToUpper ? m.gene.toUpperCase() : m.gene
+	if (!flag.data[n]) {
+		flag.data[n] = []
+	}
+	m.dt = common.dtsnvindel
+	flag.data[n].push(m)
+}

package/src/bulk.sv.js

@@ -0,0 +1,276 @@
+import * as bulk from './bulk.js'
+import * as common from './common.js'
+
+/////////////////////////////////
+//
+// client/server shared
+//
+/////////////////////////////////
+
+// work for both sv/fusion
+// must tell if the data is fusion or sv
+
+export function parseheader(line, flag, issv) {
+	const header = line.toLowerCase().split('\t')
+	if (header.length <= 1) return 'invalid file header for fusions'
+	const htry = (...lst) => {
+		for (const a of lst) {
+			const j = header.indexOf(a)
+			if (j != -1) return j
+		}
+		return -1
+	}
+	let i = htry('gene_a', 'gene1', 'genea')
+	if (i == -1) return 'gene_a missing from header'
+	header[i] = 'gene1'
+	i = htry('gene_b', 'gene2', 'geneb')
+	if (i == -1) return 'gene_b missing from header'
+	header[i] = 'gene2'
+	i = htry('chr_a', 'chr1', 'chra')
+	if (i == -1) return 'chr_a missing from header'
+	header[i] = 'chr1'
+	i = htry('chr_b', 'chr2', 'chrb')
+	if (i == -1) return 'chr_b missing from header'
+	header[i] = 'chr2'
+	i = htry('pos_a', 'position_a', 'position1', 'posa')
+	if (i == -1) return 'pos_a missing from header'
+	header[i] = 'position1'
+	i = htry('pos_b', 'position_b', 'position2', 'posb')
+	if (i == -1) return 'pos_b missing from header'
+	header[i] = 'position2'
+	i = htry('isoform_a', 'refseq_a', 'refseq1', 'isoform1', 'sv_refseqa')
+	if (i == -1) return 'isoform_a missing from header'
+	header[i] = 'isoform1'
+	i = htry('isoform_b', 'refseq_b', 'refseq2', 'isoform2', 'sv_refseqb')
+	if (i == -1) return 'isoform_b missing from header'
+	header[i] = 'isoform2'
+	i = htry('strand_a', 'orta')
+	if (i == -1) return 'strand_a missing from header'
+	header[i] = 'strand1'
+	i = htry('strand_b', 'ortb')
+	if (i == -1) return 'strand_b missing from header'
+	header[i] = 'strand2'
+	// optional
+	i = htry('sample', 'sample_name', 'tumor_sample_barcode')
+	if (i != -1) header[i] = 'sample'
+	i = htry('patient', 'donor', 'target_case_id')
+	if (i != -1) header[i] = 'patient'
+	i = htry('sampletype', 'sample type', 'sample_type')
+	if (i != -1) header[i] = 'sampletype'
+	i = htry('disease')
+	if (i != -1) header[i] = 'disease'
+	i = htry('origin')
+	if (i != -1) header[i] = 'origin'
+	if (issv) {
+		flag.sv.loaded = true
+		flag.sv.header = header
+	} else {
+		flag.fusion.loaded = true
+		flag.fusion.header = header
+	}
+	return false
+}
+
+export function parseline(i, line, flag, issv) {
+	if (line == '' || line[0] == '#') return
+	const lst = line.split('\t')
+	const m = {}
+	const header = issv ? flag.sv.header : flag.fusion.header
+	const badlines = issv ? flag.sv.badlines : flag.fusion.badlines
+
+	for (let j = 0; j < header.length; j++) {
+		m[header[j]] = lst[j]
+	}
+	if (!m.chr1) {
+		badlines.push([i, 'missing chr1', lst])
+		return
+	}
+	if (m.chr1.toLowerCase().indexOf('chr') != 0) {
+		m.chr1 = 'chr' + m.chr1
+	}
+	if (!m.chr2) {
+		badlines.push([i, 'missing chr2', lst])
+		return
+	}
+	if (m.chr2.toLowerCase().indexOf('chr') != 0) {
+		m.chr2 = 'chr' + m.chr2
+	}
+	let v = m.position1
+	if (!v) {
+		badlines.push([i, 'missing position1', lst])
+		return
+	}
+	let v2 = Number.parseInt(v)
+	if (Number.isNaN(v2) || v2 <= 0) {
+		badlines.push([i, 'invalid value for position1', lst])
+		return
+	}
+	m.position1 = v2
+	v = m.position2
+	if (!v) {
+		badlines.push([i, 'missing position2', lst])
+		return
+	}
+	v2 = Number.parseInt(v)
+	if (Number.isNaN(v2) || v2 <= 0) {
+		badlines.push([i, 'invalid value for position2', lst])
+		return
+	}
+	m.position2 = v2
+	if (bulk.parsesample(m, flag, i, lst, badlines)) {
+		return
+	}
+	if (m.isoform1 && m.isoform1.indexOf(',') != -1) {
+		const lst2 = m.isoform1.split(',')
+		m.isoform1 = undefined
+		for (const t of lst2) {
+			if (t != '') m.isoform1 = t
+		}
+	}
+	if (m.isoform2 && m.isoform2.indexOf(',') != -1) {
+		const lst2 = m.isoform2.split(',')
+		m.isoform2 = undefined
+		for (const t of lst2) {
+			if (t != '') m.isoform2 = t
+		}
+	}
+	if (!m.gene1) {
+		m.isoform1 = undefined
+	}
+	if (!m.gene2) {
+		m.isoform2 = undefined
+	}
+	if (m.gene1) {
+		// put data under gene1
+		flag.good++
+		const m2 = {
+			dt: issv ? common.dtsv : common.dtfusionrna,
+			class: issv ? common.mclasssv : common.mclassfusionrna,
+			isoform: m.isoform1,
+			mname: m.gene2 || m.chr2,
+			sample: m.sample,
+			patient: m.patient,
+			sampletype: m.sampletype,
+			origin: m.origin,
+			disease: m.disease,
+			pairlst: [
+				{
+					a: {
+						name: m.gene1,
+						isoform: m.isoform1,
+						strand: m.strand1,
+						chr: m.chr1,
+						position: m.position1
+					},
+					b: {
+						name: m.gene2,
+						isoform: m.isoform2,
+						strand: m.strand2,
+						chr: m.chr2,
+						position: m.position2
+					}
+				}
+			]
+		}
+		const n = flag.geneToUpper ? m.gene1.toUpperCase() : m.gene1
+		if (!flag.data[n]) {
+			flag.data[n] = []
+		}
+		flag.data[n].push(m2)
+	}
+	if (m.gene2 && m.gene2 != m.gene1) {
+		// put data under gene2
+		flag.good++
+		const m2 = {
+			dt: issv ? common.dtsv : common.dtfusionrna,
+			class: issv ? common.mclasssv : common.mclassfusionrna,
+			isoform: m.isoform2,
+			mname: m.gene1 || m.chr1,
+			sample: m.sample,
+			patient: m.patient,
+			sampletype: m.sampletype,
+			origin: m.origin,
+			disease: m.disease,
+			pairlst: [
+				{
+					a: {
+						name: m.gene1,
+						isoform: m.isoform1,
+						strand: m.strand1,
+						chr: m.chr1,
+						position: m.position1
+					},
+					b: {
+						name: m.gene2,
+						isoform: m.isoform2,
+						strand: m.strand2,
+						chr: m.chr2,
+						position: m.position2
+					}
+				}
+			]
+		}
+		const n = flag.geneToUpper ? m.gene2.toUpperCase() : m.gene2
+		if (!flag.data[n]) {
+			flag.data[n] = []
+		}
+		flag.data[n].push(m2)
+	}
+}
+
+export function duplicate(m) {
+	const n = {}
+	for (const k in m) {
+		if (k == 'pairlst') continue
+		const v = m[k]
+		const type = typeof v
+		if (type == 'object') {
+			continue
+		}
+		n[k] = v
+	}
+	if (m.pairlst) {
+		n.pairlst = []
+		for (const pair of m.pairlst) {
+			const p = {}
+			for (const k in pair) {
+				if (k == 'a' || k == 'b' || k == 'interstitial') {
+					continue
+				}
+				p[k] = pair[k]
+			}
+			if (pair.a) {
+				p.a = {}
+				for (const k in pair.a) {
+					const v = pair.a[k]
+					if (typeof v == 'object') {
+						continue
+					}
+					p.a[k] = v
+				}
+			}
+			if (pair.b) {
+				p.b = {}
+				for (const k in pair.b) {
+					const v = pair.b[k]
+					if (typeof v == 'object') {
+						continue
+					}
+					p.b[k] = v
+				}
+			}
+			if (pair.interstitial) {
+				p.interstitial = {}
+				for (const k in pair.interstitial) {
+					const v = pair.interstitial[k]
+					if (typeof v == 'object') {
+						continue
+					}
+					p.interstitial[k] = v
+				}
+			}
+			n.pairlst.push(p)
+		}
+	}
+	return n
+}

package/src/bulk.svjson.js

@@ -0,0 +1,162 @@
+/////////////////////////////////
+//
+// client/server shared
+//
+/////////////////////////////////
+
+import * as common from './common.js'
+import * as bulk from './bulk.js'
+
+// work for both sv/fusion
+// must tell if the data is fusion or sv
+
+export function parseheader(line, flag) {
+	const header = line.toLowerCase().split('\t')
+	if (header.length <= 1) return 'invalid file header for svjson'
+	const htry = (...lst) => {
+		for (const a of lst) {
+			const j = header.indexOf(a)
+			if (j != -1) return j
+		}
+		return -1
+	}
+	let i = htry('sample')
+	if (i != -1) header[i] = 'sample'
+	i = htry('sampletype')
+	if (i != -1) header[i] = 'sampletype'
+	i = htry('patient')
+	if (i != -1) header[i] = 'patient'
+	i = htry('json', 'jsontext')
+	if (i == -1) return ['json missing from header']
+	header[i] = 'jsontext'
+	return [null, header]
+}
+
+export function parseline(i, line, flag, header) {
+	if (line == '' || line[0] == '#') return
+	const lst = line.split('\t')
+	const m = {}
+	const badlines = flag.svjson.badlines
+
+	for (let j = 0; j < header.length; j++) {
+		m[header[j]] = lst[j]
+	}
+	if (!m.jsontext) {
+		badlines.push([i, 'missing jsontext', lst])
+		return
+	}
+	if (bulk.parsesample(m, flag, i, lst, badlines)) {
+		return
+	}
+	let json
+	try {
+		json = JSON.parse(m.jsontext)
+	} catch (e) {
+		badlines.push([i, 'invalid JSON text', lst])
+		return
+	}
+	// duplicating logic in pediatric.js
+	if (Array.isArray(json)) {
+		// json is pairlst
+		for (const pair of json) {
+			if (pair.a && pair.a.name && pair.a.isoform) {
+				flag.good++
+				const m2 = {
+					dt: common.dtfusionrna,
+					class: common.mclassfusionrna,
+					isoform: pair.a.isoform,
+					mname: pair.b.name
+				}
+				for (const k in m) {
+					if (k != 'jsontext') m2[k] = m[k]
+				}
+				m2.pairlst = duplicate(json)
+				const n = pair.a.name.toUpperCase()
+				if (!flag.data[n]) {
+					flag.data[n] = []
+				}
+				flag.data[n].push(m2)
+			}
+			if (pair.b && pair.b.name && pair.b.isoform) {
+				flag.good++
+				const m2 = {
+					dt: common.dtfusionrna,
+					class: common.mclassfusionrna,
+					isoform: pair.b.isoform,
+					mname: pair.a.name
+				}
+				for (const k in m) {
+					if (k != 'jsontext') m2[k] = m[k]
+				}
+				m2.pairlst = duplicate(json)
+				const n = pair.b.name.toUpperCase()
+				if (!flag.data[n]) {
+					flag.data[n] = []
+				}
+				flag.data[n].push(m2)
+			}
+		}
+	} else {
+		json.dt = json.typecode
+		delete json.typecode
+		switch (json.dt) {
+			case common.dtitd:
+				json.class = common.mclassitd
+				json.mname = 'ITD'
+				break
+			case common.dtnloss:
+				json.class = common.mclassnloss
+				json.mname = 'N-loss'
+				break
+			case common.dtcloss:
+				json.class = common.mclasscloss
+				json.mname = 'C-loss'
+				break
+			case common.dtdel:
+				json.class = common.mclassdel
+				json.mname = 'Del'
+				break
+			case common.dtsv:
+				json.class = common.mclasssv
+				json.mname = 'SV'
+				break
+			default:
+				badlines.push([i, 'unknown datatype', lst])
+				return
+		}
+		// record only about a single gene
+		if (!json.gene) {
+			badlines.push([i, 'json.gene missing', lst])
+			return
+		}
+		flag.good++
+		for (const k in m) {
+			if (k != 'jsontext') {
+				json[k] = m[k]
+			}
+		}
+		const n = flag.geneToUpper ? json.gene.toUpperCase() : json.gene.toUpperCase()
+		if (!flag.data[n]) {
+			flag.data[n] = []
+		}
+		flag.data[n].push(json)
+	}
+}
+
+function duplicate(lst) {
+	const d = []
+	for (const pair of lst) {
+		const p = { a: {}, b: {} }
+		for (const k in pair) {
+			if (k != 'a' && k != 'b') p[k] = pair[k]
+		}
+		for (const k in pair.a) {
+			p.a[k] = pair.a[k]
+		}
+		for (const k in pair.b) {
+			p.b[k] = pair.b[k]
+		}
+		d.push(p)
+	}
+	return d
+}