Merge pull request #19 from bento-platform/releases/v3.2

releases/v3.2

README.md

@@ -204,7 +204,7 @@ Response
 Requests
 > &nbsp;&nbsp;**GET** `/variants/get/by/variantId`<br/>
 > &nbsp;&nbsp;&nbsp;params: 
->   - chromosome : **string** `( 1-23, X, Y, M )`
+>   - chromosome : **string** `( 1-23, X, Y, MT )`
 >   - lowerBound : **number**
 >   - upperBound : **number**
 >   - reference : **string** `an allele ( "A" | "C" | "G" | "T"  or some combination thereof )`
@@ -214,10 +214,11 @@ Requests
 >   - sortByPosition : **string** `(<empty> | asc | desc)`
 >   - includeInfoInResultSet : **boolean** `(true | false)`
 >   - genotype : **string** `( "HETEROZYGOUS" | "HOMOZYGOUS_REFERENCE" | "HOMOZYGOUS_ALTERNATE" )`
+>   - getSampleIdsOnly : **bool**  *`(optional) -  default: false  `*
 >
 > &nbsp;&nbsp;**GET** `/variants/count/by/variantId`<br/>
 > &nbsp;&nbsp;&nbsp;params: 
->   - chromosome : **string** `( 1-23, X, Y, M )`
+>   - chromosome : **string** `( 1-23, X, Y, MT )`
 >   - lowerBound : **number**
 >   - upperBound : **number**
 >   - reference : **string** `an allele`
@@ -227,7 +228,7 @@ Requests
 
 > &nbsp;&nbsp;**GET** `/variants/get/by/sampleId`<br/>
 > &nbsp;&nbsp;&nbsp;params: 
->   - chromosome : **string** `( 1-23, X, Y, M )`
+>   - chromosome : **string** `( 1-23, X, Y, MT )`
 >   - lowerBound : **number**
 >   - upperBound : **number**
 >   - reference : **string** `an allele`
@@ -240,7 +241,7 @@ Requests
 >
 > &nbsp;&nbsp;**GET** `/variants/count/by/sampleId`<br/>
 > &nbsp;&nbsp;&nbsp;params: 
->   - chromosome : **string** `( 1-23, X, Y, M )`
+>   - chromosome : **string** `( 1-23, X, Y, MT )`
 >   - lowerBound : **number**
 >   - upperBound : **number**
 >   - reference : **string** `an allele`
@@ -257,14 +258,18 @@ Generalized Response Body Structure
 >{
 >     "status":  `number` (200 - 500),
 >     "message": `string` ("Success" | "Error"),
->     "data": [
+>     "results": [
 >         {
->             "variantId":  `string`,
->             "sampleId":  `string`,
->             "count":  `number`,
->             "results": [
+>             "query":  `string`,       // reflective of the type of id queried for, i.e 'variantId:abc123', or 'sampleId:HG0001
+>             "assemblyId": `string` ("GRCh38" | "GRCh37" | "NCBI36" | "Other"),    // reflective of the assembly id queried for
+>             "count":  `number`,   // this field is only present when performing a COUNT query
+>             "start":  `number`,   // reflective of the provided lowerBound parameter, 0 if none
+>             "end":  `number`,     // reflective of the provided upperBound parameter, 0 if none
+>             "chromosome":  `string`,       // reflective of the chromosome queried for
+>             "calls": [            // this field is only present when performing a GET query
 >                 {
->                    "filter": `string`,
+>                    "id": `string`, // variantId
+>                    "chrom":  `string`,
 >                    "pos": `number`,
 >                    "ref": `[]string`,  // list of alleles
 >                    "alt": `[]string`,  // list of alleles
@@ -277,21 +282,9 @@ Generalized Response Body Structure
 >                    ],
 >                    "format":`string`,
 >                    "qual": `number`,
->                    "id": `string`,
->                    "sample": {
->                        "id": `string`,
->                        "variation": {
->                            "genotype": {
->                                "phased": `boolean`,
->                                "alleleLeft": `number`,
->                                "alleleRight": `number`,
->                                "zygosity": `number` (0 : "Unknown" | 1 : "Heterozygous" | 2 : "HomozygousReference" | 3 : "HomozygousAlternate")
->                            },
->                            "genotypeProbability": `[]float` | null,
->                            "phredScaleLikelyhood": `[]float` | null,
->                        }
->                    },
->                    "fileId": `string` (UUID),
+>                    "filter": `string`,
+>                    "sampleId": `string`,
+>                    "genotype_type": `string ( "HETEROZYGOUS" | "HOMOZYGOUS_REFERENCE" | "HOMOZYGOUS_ALTERNATE" )`,
 >                    "assemblyId": `string` ("GRCh38" | "GRCh37" | "NCBI36" | "Other"),
 >                 },
 >                 ...

docker-compose.yaml

@@ -48,6 +48,7 @@ services:
       - GOHAN_API_VCF_PATH=${GOHAN_API_CONTAINERIZED_VCF_PATH}
       - GOHAN_API_GTF_PATH=${GOHAN_API_CONTAINERIZED_GTF_PATH}
       - GOHAN_API_API_DRS_BRIDGE_DIR=${GOHAN_API_API_DRS_BRIDGE_DIR_CONTAINERIZED}
+      - GOHAN_API_BULK_INDEXING_CAP=${GOHAN_API_BULK_INDEXING_CAP}
       - GOHAN_API_FILE_PROC_CONC_LVL=${GOHAN_API_FILE_PROC_CONC_LVL}
       - GOHAN_API_LINE_PROC_CONC_LVL=${GOHAN_API_LINE_PROC_CONC_LVL}
 

src/api/models/constants/main.go

@@ -1,5 +1,7 @@
 package constants
 
+var VcfHeaders = []string{"chrom", "pos", "id", "ref", "alt", "qual", "filter", "info", "format"}
+
 /*
 	Defines a set of base level
 	constants and enums to be used

src/api/models/constants/zygosity/main.go

@@ -14,3 +14,18 @@ const (
 func IsKnown(value int) bool {
 	return value > int(Unknown) && value <= int(HomozygousAlternate)
 }
+
+func ZygosityToString(zyg constants.Zygosity) string {
+	switch zyg {
+	case Heterozygous:
+		return "HETEROZYGOUS"
+	case HomozygousAlternate:
+		return "HOMOZYGOUS_ALTERNATE"
+	case HomozygousReference:
+		return "HOMOZYGOUS_REFERENCE"
+	default:
+		return "UNKNOWN"
+	}
+}
+
+// TODO: StringOrIntToZygosity

src/api/models/dtos/main.go

@@ -0,0 +1,65 @@
+package dtos
+
+import (
+	"api/models/constants"
+	"api/models/indexes"
+)
+
+// ---- Variants
+type VariantReponse struct {
+	Status  int    `json:"status,omitempty"`
+	Message string `json:"message,omitempty"`
+}
+type VariantGetReponse struct {
+	VariantReponse
+	Results []VariantGetResult `json:"results"`
+}
+type VariantCountReponse struct {
+	VariantReponse
+	Results []VariantCountResult `json:"results"`
+}
+
+type VariantResult struct {
+	Query      string               `json:"query,omitempty"`
+	AssemblyId constants.AssemblyId `json:"assembly_id"`
+	Chromosome string               `json:"chromosome"`
+	Start      int                  `json:"start"`
+	End        int                  `json:"end"`
+}
+
+type VariantGetResult struct {
+	VariantResult
+	Calls []VariantCall `json:"calls"`
+}
+type VariantCountResult struct {
+	VariantResult
+	Count int `json:"count"`
+}
+
+type VariantCall struct {
+	Chrom  string   `json:"chrom,omitempty"`
+	Pos    int      `json:"pos,omitempty"`
+	Id     string   `json:"id,omitempty"`
+	Ref    []string `json:"ref,omitempty"`
+	Alt    []string `json:"alt,omitempty"`
+	Format []string `json:"format,omitempty"`
+	Qual   int      `json:"qual,omitempty"`
+	Filter string   `json:"filter,omitempty"`
+
+	Info []indexes.Info `json:"info,omitempty"` // TODO; refactor?
+
+	SampleId     string `json:"sample_id"`
+	GenotypeType string `json:"genotype_type"`
+	// TODO: GenotypeProbability, PhredScaleLikelyhood ?
+
+	AssemblyId constants.AssemblyId `json:"assemblyId,omitempty"`
+}
+
+// -- Genes
+type GenesResponseDTO struct {
+	Status  int            `json:"status"`
+	Message string         `json:"message"`
+	Term    string         `json:"term"`
+	Count   int            `json:"count"`
+	Results []indexes.Gene `json:"results"` // []Gene
+}

src/api/models/elasticsearch.go → src/api/models/indexes/main.go

@@ -1,11 +1,9 @@
-package models
+package indexes
 
 import (
 	c "api/models/constants"
 )
 
-var VcfHeaders = []string{"chrom", "pos", "id", "ref", "alt", "qual", "filter", "info", "format"}
-
 type Variant struct {
 	Chrom  string   `json:"chrom"`
 	Pos    int      `json:"pos"`

src/api/models/ingest/structs/main.go

@@ -1,16 +1,16 @@
 package structs
 
 import (
-	"api/models"
+	"api/models/indexes"
 	"sync"
 )
 
 type IngestionQueueStructure struct {
-	Variant   *models.Variant
+	Variant   *indexes.Variant
 	WaitGroup *sync.WaitGroup
 }
 
 type GeneIngestionQueueStructure struct {
-	Gene      *models.Gene
+	Gene      *indexes.Gene
 	WaitGroup *sync.WaitGroup
 }

src/api/mvc/genes.go

@@ -2,10 +2,10 @@ package mvc
 
 import (
 	"api/contexts"
-	"api/models"
 	"api/models/constants"
 	assemblyId "api/models/constants/assembly-id"
 	"api/models/constants/chromosome"
+	"api/models/dtos"
 	"api/models/ingest"
 	"api/models/ingest/structs"
 	esRepo "api/repositories/elasticsearch"
@@ -22,6 +22,8 @@ import (
 	"sync"
 	"time"
 
+	"api/models/indexes"
+
 	"github.com/labstack/echo"
 	"github.com/mitchellh/mapstructure"
 )
@@ -268,7 +270,7 @@ func GenesIngest(c echo.Context) error {
 							return
 						}
 
-						discoveredGene := &models.Gene{
+						discoveredGene := &indexes.Gene{
 							Name:       geneName,
 							Chrom:      chromosomeClean,
 							Start:      start,
@@ -370,13 +372,13 @@ func GenesGetByNomenclatureWildcard(c echo.Context) error {
 	mapstructure.Decode(docsHits, &allDocHits)
 
 	// grab _source for each hit
-	var allSources []models.Gene
+	var allSources []indexes.Gene
 
 	for _, r := range allDocHits {
 		source := r["_source"].(map[string]interface{})
 
 		// cast map[string]interface{} to struct
-		var resultingVariant models.Gene
+		var resultingVariant indexes.Gene
 		mapstructure.Decode(source, &resultingVariant)
 
 		// accumulate structs
@@ -385,7 +387,7 @@ func GenesGetByNomenclatureWildcard(c echo.Context) error {
 
 	fmt.Printf("Found %d docs!\n", len(allSources))
 
-	geneResponseDTO := models.GenesResponseDTO{
+	geneResponseDTO := dtos.GenesResponseDTO{
 		Term:    term,
 		Count:   len(allSources),
 		Results: allSources,