seekia/resources/geneticReferences/geneticReferences_test.go

// verifyGeneticReferences provides functions to run a check to make sure the genetic resources are valid and have no conflicts

package verifyGeneticReferences

// We check to make sure:
//	1. No identifier collisions exist
//	2. No disease/trait name collisions exist
//	4. Verifies the minimum and maximum risk weights for each polygenic disease locus
//	5. Each identifier is the correct format (3 bytes encoded hex)

// Identifiers are 3 bytes/24 bits long, so there is at least a 1 in 16 million chance that two will collide when generating them randomly

import "seekia/resources/geneticReferences/locusMetadata"
import "seekia/resources/geneticReferences/monogenicDiseases"
import "seekia/resources/geneticReferences/polygenicDiseases"
import "seekia/resources/geneticReferences/traits"

import "seekia/internal/helpers"
import "seekia/internal/encoding"

import "testing"
import "strings"
import "slices"

func TestGeneticReferences(t *testing.T){

	verifyIdentifier := func(inputIdentifier string)bool{

		_, err := encoding.DecodeHexStringTo3ByteArray(inputIdentifier)
		if (err != nil) {
			return false
		}

		return true
	}

	verifyBase := func(inputBase string)bool{

		if (inputBase != "A" && inputBase != "G" && inputBase != "C" && inputBase != "T" && inputBase != "I" && inputBase != "D"){
			return false
		}
		return true
	}

	verifyBasePair := func(inputBasePair string)bool{

		baseA, baseB, delimiterFound := strings.Cut(inputBasePair, ";")
		if (delimiterFound == false){
			return false
		}

		baseIsValid := verifyBase(baseA)
		if (baseIsValid == false){
			return false
		}

		baseIsValid = verifyBase(baseB)
		if (baseIsValid == false){
			return false
		}
		return true
	}

	verifyReferencesMap := func(inputReferencesMap map[string]string)bool{

		if (len(inputReferencesMap) == 0){
			return true
		}

		for referenceName, referenceLink := range inputReferencesMap{

			if (referenceName == ""){
				return false
			}
			if (referenceLink == ""){
				return false
			}
		}

		return true
	}

	monogenicDiseases.InitializeMonogenicDiseaseVariables()

	monogenicDiseasesObjectsList, err := monogenicDiseases.GetMonogenicDiseaseObjectsList()
	if (err != nil){
		t.Fatalf("Failed to get monogenic disease objects list: " + err.Error())
	}

	// We use this map to make sure all RSIDs have metadata in locusMetadata
	allRSIDsMap := make(map[int64]struct{})

	allIdentifiersMap := make(map[string]struct{})

	monogenicDiseaseNamesMap := make(map[string]struct{})

	for _, diseaseObject := range monogenicDiseasesObjectsList{

		diseaseName := diseaseObject.DiseaseName
		diseaseGeneName := diseaseObject.GeneName
		dominantOrRecessive := diseaseObject.DominantOrRecessive
		variantsList := diseaseObject.VariantsList
		diseaseReferencesMap := diseaseObject.References

		if (diseaseName == ""){
			t.Fatalf("Monogenic Disease name is empty.")
		}
		_, exists := monogenicDiseaseNamesMap[diseaseName]
		if (exists == true){
			t.Fatalf("Monogenic Disease name collision found: " + diseaseName)
		}
		monogenicDiseaseNamesMap[diseaseName] = struct{}{}

		// Monogenic disease names cannot contain underscores
		// This is because when we encode monogenic disease names in user profiles, we replace the whitespace with underscores
		// We have to be able to reliably undo this
		containsUnderscore := strings.Contains(diseaseName, "_")
		if (containsUnderscore == true){
			t.Fatalf("Monogenic Disease name contains underscore: " + diseaseName)
		}

		if (diseaseGeneName == ""){
			t.Fatalf("Monogenic Disease gene name is empty: " + diseaseName)
		}
		if (dominantOrRecessive	!= "Dominant" && dominantOrRecessive != "Recessive"){
			t.Fatalf("Monogenic Disease dominantOrRecessive is invalid: " + diseaseName)
		}

		referencesAreValid := verifyReferencesMap(diseaseReferencesMap)
		if (referencesAreValid == false){
			t.Fatalf("Monogenic Disease references are invalid: " + diseaseName)
		}

		if (len(variantsList) == 0){
			t.Fatalf("Monogenic Disease contains no variants: " + diseaseName) 		
		}

		for _, variantObject := range variantsList{

			variantIdentifier := variantObject.VariantIdentifier
			variantRSID := variantObject.VariantRSID
			variantNamesList := variantObject.VariantNames
			variantHealthyBase := variantObject.HealthyBase
			variantDefectiveBase := variantObject.DefectiveBase
			variantReferences := variantObject.References

			allRSIDsMap[variantRSID] = struct{}{}

			identifierIsValid := verifyIdentifier(variantIdentifier)
			if (identifierIsValid == false){
				t.Fatalf(diseaseName + " Invalid variant identifier found: " + variantIdentifier)
			}

			_, exists := allIdentifiersMap[variantIdentifier]
			if (exists == true){
				t.Fatalf(diseaseName + " Duplicate variant identifier found: " + variantIdentifier)
			}
			allIdentifiersMap[variantIdentifier] = struct{}{}

			if (len(variantNamesList) == 0){
				t.Fatalf("Variant names list is empty: " + variantIdentifier)
			}
			for _, variantName := range variantNamesList{
				if (variantName == ""){
					t.Fatalf("Variant name is empty: " + variantIdentifier)
				}	
			}

			healthyBaseIsValid := verifyBase(variantHealthyBase)
			defectiveBaseIsValid := verifyBase(variantDefectiveBase)

			if (healthyBaseIsValid == false || defectiveBaseIsValid == false){
				t.Fatalf(diseaseName + " Invalid healthy/defective base found: " + variantIdentifier)
			}

			if (variantHealthyBase == variantDefectiveBase){
				t.Fatalf(diseaseName + " Identical healthy/defective bases found: " + variantIdentifier)
			}

			referencesAreValid := verifyReferencesMap(variantReferences)
			if (referencesAreValid == false){
				t.Fatalf("Disease variant references map is invalid: " + variantIdentifier)
			}
		}
	}

	err = polygenicDiseases.InitializePolygenicDiseaseVariables()
	if (err != nil){
		t.Fatalf("InitializePolygenicDiseaseVariables failed: " + err.Error())
	}

	polygenicDiseaseObjectsList, err := polygenicDiseases.GetPolygenicDiseaseObjectsList()
	if (err != nil) {
		t.Fatalf("Failed to get polygenicDisease objects list: " + err.Error())
	}

	polygenicDiseaseNamesMap := make(map[string]struct{})

	for _, diseaseObject := range polygenicDiseaseObjectsList{

		diseaseName := diseaseObject.DiseaseName
		diseaseDescription := diseaseObject.DiseaseDescription
		diseaseEffectedSex := diseaseObject.EffectedSex
		diseaseLocusReferencesMap := diseaseObject.LocusReferencesMap
		diseaseLociList := diseaseObject.LociList
		diseaseReferencesMap := diseaseObject.References

		if (diseaseName == ""){
			t.Fatalf("PolygenicDisease name is empty.")
		}
		_, exists := polygenicDiseaseNamesMap[diseaseName]
		if (exists == true){
			t.Fatalf("PolygenicDisease name collision found: " + diseaseName)
		}
		polygenicDiseaseNamesMap[diseaseName] = struct{}{}

		if (diseaseDescription == ""){
			t.Fatalf("PolygenicDisease description is empty for disease: " + diseaseName)
		}
		if (diseaseEffectedSex != "Male" && diseaseEffectedSex != "Female" && diseaseEffectedSex != "Both"){
			t.Fatalf("PolygenicDisease effected sex is invalid: " + diseaseEffectedSex)
		}

		for rsID, referencesMap := range diseaseLocusReferencesMap{

			containsItem := slices.Contains(diseaseLociList, rsID)
			if (containsItem == false){
				t.Fatalf("Polygenic disease diseaseLocusReferencesMap contains disease locus that is not inside of the disease's loci list.")
			}

			allRSIDsMap[rsID] = struct{}{}

			referencesAreValid := verifyReferencesMap(referencesMap)
			if (referencesAreValid == false){
				t.Fatalf("PolygenicDisease references map is invalid for disease locus.")
			}
		}

		containsDuplicates, _ := helpers.CheckIfListContainsDuplicates(diseaseLociList)
		if (containsDuplicates == true){
			t.Fatalf("Polygenic disease object contains diseaseLociList with duplicate rsIDs.")
		}

		if (len(diseaseLocusReferencesMap) > len(diseaseLociList)){
			t.Fatalf("Polygenic disease contains locus references map that is longer than the diseaseLociList")
		}

		referencesAreValid := verifyReferencesMap(diseaseReferencesMap)
		if (referencesAreValid == false){
			t.Fatalf("PolygenicDisease references map is invalid for disease: " + diseaseName)
		}
	}

	err = traits.InitializeTraitVariables()
	if (err != nil){
		t.Fatalf("InitializeTraitVariables failed: " + err.Error())
	}

	traitObjectsList, err := traits.GetTraitObjectsList()
	if (err != nil){
		t.Fatalf("Failed to get trait objects list: " + err.Error())
	}

	traitNamesMap := make(map[string]struct{})

	for _, traitObject := range traitObjectsList{

		traitName := traitObject.TraitName
		traitDescription := traitObject.TraitDescription
		traitDiscreteOrNumeric := traitObject.DiscreteOrNumeric
		traitLocusReferencesMap := traitObject.LocusReferencesMap
		traitLociList := traitObject.LociList
		traitLociList_Rules := traitObject.LociList_Rules
		traitRulesList := traitObject.RulesList
		traitOutcomesList := traitObject.OutcomesList
		traitReferencesMap := traitObject.ReferencesMap

		if (traitName == ""){
			t.Fatalf("Empty trait name exists.")
		}
		_, exists := traitNamesMap[traitName]
		if (exists == true){
			t.Fatalf("Duplicate trait name exists: " + traitName)
		}
		traitNamesMap[traitName] = struct{}{}

		if (traitDescription == ""){
			t.Fatalf("Empty trait description exists for trait: " + traitName)
		}
		if (traitDiscreteOrNumeric != "Discrete" && traitDiscreteOrNumeric != "Numeric"){
			t.Fatalf("Invalid DiscreteOrNumeric for trait: " + traitDiscreteOrNumeric)
		}
		if (len(traitOutcomesList) != 0){

			if (len(traitOutcomesList) < 2){
				t.Fatalf("Not enough trait outcomes for trait: " + traitName)
			}
			for _, traitOutcome := range traitOutcomesList{
				if (traitOutcome == ""){
					t.Fatalf("Empty trait outcome exists for trait: " + traitName)
				}
			}
		} else {

			// If there are no outcomes, then no rules can exist
			if (len(traitRulesList) != 0){
				t.Fatalf("Trait outcomes list is empty, trait rules list is not.")
			}
		}

		referencesAreValid := verifyReferencesMap(traitReferencesMap)
		if (referencesAreValid == false){
			t.Fatalf("Invalid references exist for trait: " + traitName)
		}

		if (len(traitLocusReferencesMap) == 0){
			t.Fatalf("No trait locus references exist for trait: " + traitName)
		}

		for locusRSID, locusReferences := range traitLocusReferencesMap{

			allRSIDsMap[locusRSID] = struct{}{}

			if (locusReferences == nil){
				t.Fatalf("A trait locus has no references map: " + traitName)
			}
			if (len(locusReferences) == 0){
				t.Fatalf("A trait locus has no references: " + traitName)
			}

			locusExists := slices.Contains(traitLociList, locusRSID)
			if (locusExists == false){
				t.Fatalf("traitLocusReferencesMap contains rsID which does not exist in traitLociList")
			}
		}

		if (len(traitLociList) == 0){
			t.Fatalf("No trait loci exist for trait: " + traitName)
		}

		for _, rsID := range traitLociList{
			allRSIDsMap[rsID] = struct{}{}
		}

		for _, rsID := range traitLociList_Rules{

			locusExists := slices.Contains(traitLociList, rsID)
			if (locusExists == false){
				t.Fatalf("traitLociList_Rules contains locus not present in traitLociList")
			}
		}

		if (len(traitRulesList) == 0){
			// No rules exist.
			continue
		}

		for _, ruleObject := range traitRulesList{

			ruleIdentifier := ruleObject.RuleIdentifier
			ruleLociList := ruleObject.LociList
			ruleOutcomePointsMap := ruleObject.OutcomePointsMap
			ruleReferencesMap := ruleObject.ReferencesMap

			identifierIsValid := verifyIdentifier(ruleIdentifier)
			if (identifierIsValid == false){
				t.Fatalf("Invalid identifier exists: " + ruleIdentifier)
			}
			_, exists := allIdentifiersMap[ruleIdentifier]
			if (exists == true){
				t.Fatalf("Duplicate identifier exists: " + ruleIdentifier)
			}
			allIdentifiersMap[ruleIdentifier] = struct{}{}

			if (len(ruleOutcomePointsMap) == 0){
				t.Fatalf("Rule contains empty rule outcome points map: " + ruleIdentifier)
			}

			for outcomeName, _ := range ruleOutcomePointsMap{
				isValid := slices.Contains(traitOutcomesList, outcomeName)
				if (isValid == false){
					t.Fatalf("Rule outcome points map contains invalid outcome: " + outcomeName)
				}
			}

			if (len(ruleLociList) == 0){
				t.Fatalf("Rule contains empty rule loci list: " + ruleIdentifier)
			}

			for _, locusObject := range ruleLociList{

				locusIdentifier := locusObject.LocusIdentifier
				locusRSID := locusObject.LocusRSID
				locusBasePairsList := locusObject.BasePairsList

				allRSIDsMap[locusRSID] = struct{}{}

				isValid := verifyIdentifier(locusIdentifier)
				if (isValid == false){
					t.Fatalf("Trait rule Locus identifier is invalid: " + locusIdentifier)
				}

				_, mapContainsItem := traitLocusReferencesMap[locusRSID]
				if (mapContainsItem == false){
					t.Fatalf("Rule locus contains rsid which is not contained within LocusReferencesMap.")
				}

				sliceContainsItem := slices.Contains(traitLociList, locusRSID)
				if (sliceContainsItem == false){
					t.Fatalf("Rule locus contains rsid which is not contained within traitLociList.")
				}

				sliceContainsItem = slices.Contains(traitLociList_Rules, locusRSID)
				if (sliceContainsItem == false){
					t.Fatalf("Rule locus contains rsid which is not contained within traitLociList_Rules.")
				}

				if (len(locusBasePairsList) == 0){
					t.Fatalf("Trait rule locus base pairs list is empty: " + locusIdentifier)
				}
				for _, locusBasePair := range locusBasePairsList{

					basePairIsValid := verifyBasePair(locusBasePair)
					if (basePairIsValid == false){
						t.Fatalf("Rule Locus base pairs list contains invalid base pair: " + locusBasePair)
					}
				}
			}

			referencesAreValid := verifyReferencesMap(ruleReferencesMap)
			if (referencesAreValid == false){
				t.Fatalf("Invalid references map for trait rule locus: " + ruleIdentifier)
			}
		}
	}

	err = locusMetadata.InitializeLocusMetadataVariables()
	if (err != nil){
		t.Fatalf("Failed to initialize locus metadata variables: " + err.Error())
	}

	locusMetadataObjectsList, err := locusMetadata.GetLocusMetadataObjectsList()
	if (err != nil){
		t.Fatalf("GetLocusMetadataObjectsList failed: " + err.Error())
	}

	// We use the locusPositionsMap to make sure there are no locations that refer to the same position on the same chromosome

	type locusPositionStruct struct{
		chromosome int
		position int
	}

	locusPositionsMap := make(map[locusPositionStruct]struct{})

	// We use the companyAliasesMap to make sure there are no company alias collisions.
	//
	// We only care about alias collisions within each company.
	// Multiple companies can refer to the same location with the same alias.

	type companyAliasStruct struct{

		geneticsCompany locusMetadata.GeneticsCompany
		locusAlias string
	}

	companyAliasesMap := make(map[companyAliasStruct]struct{})

	// We use this map to make sure that locus metadata rsIDs do not collide.
	// We don't want any duplicate rsIDs within any of the loci.
	locusMetadataRSIDsMap := make(map[int64]struct{})

	for _, locusMetadataObject := range locusMetadataObjectsList{

		rsidsList := locusMetadataObject.RSIDsList
		locusChromosome := locusMetadataObject.Chromosome
		locusPosition := locusMetadataObject.Position
		geneInfoIsKnown := locusMetadataObject.GeneInfoIsKnown
		geneExists := locusMetadataObject.GeneExists
		geneNamesList := locusMetadataObject.GeneNamesList
		locusCompanyAliasesMap := locusMetadataObject.CompanyAliases
		referencesMap := locusMetadataObject.References

		if (len(rsidsList) == 0){
			t.Fatalf("locusMetadataObjectsList contains locus with empty RSIDs list.")
		}

		// The primary RSID is the only rsID which should appear in the genetic references
		// The primary RSID is the first rsID in the locus rsIDs list
		primaryRSID := rsidsList[0]

		_, exists := allRSIDsMap[primaryRSID]
		if (exists == false){
			t.Fatalf("locusMetadataObjectsList contains unnecessary locus: No matching rsids exist.")
		}

		for index, rsID := range rsidsList{

			_, exists := locusMetadataRSIDsMap[rsID]
			if (exists == true){

				rsidString := helpers.ConvertInt64ToString(rsID)
				t.Fatalf("locusMetadataObjectsList contains duplicate RSID: " + rsidString)
			}

			locusMetadataRSIDsMap[rsID] = struct{}{}

			if (index != 0){

				// This is not a primary rsID
				_, exists = allRSIDsMap[rsID]
				if (exists == true){
					rsIDString := helpers.ConvertInt64ToString(rsID)
					t.Fatalf("allRSIDsMap contains non-primary rsID: " + rsIDString)
				}			
			}
		}

		if (locusChromosome == 0){
			// 0 is uninitialized.
			t.Fatalf("locusMetadataObjectsList contains locus with 0 chromosome.")
		}

		if (locusPosition == 0){
			// 0 is uninitialized.
			t.Fatalf("locusMetadataObjectsList contains locus with 0 position.")
		}

		locusPositionObject := locusPositionStruct{
			chromosome: locusChromosome,
			position: locusPosition,
		}

		_, exists = locusPositionsMap[locusPositionObject]
		if (exists == true){
			t.Fatalf("locusMetadataObjectsList contains locus position collision.")
		}

		locusPositionsMap[locusPositionObject] = struct{}{}

		if (geneInfoIsKnown == true && geneExists == true){

			if (len(geneNamesList) == 0){
				t.Fatalf("locusMetadataObjectsList contains locus with known gene and empty geneNamesList.")
			}

			for _, geneName := range geneNamesList{
				if (geneName == ""){
					t.Fatalf("locusMetadataObjectsList contains locus with empty geneName in geneNamesList.")
				}
			}
		}

		for companyObject, companyAliasesList := range locusCompanyAliasesMap{

			for _, locusCompanyAlias := range companyAliasesList{

				companyAliasObject := companyAliasStruct{
					geneticsCompany: companyObject,
					locusAlias: locusCompanyAlias,
				}

				_, exists := companyAliasesMap[companyAliasObject]
				if (exists == true){
					t.Fatalf("locusMetadataObjectsList contains companyAlias collision: " + locusCompanyAlias)
				}

				companyAliasesMap[companyAliasObject] = struct{}{}
			}
		}

		isValid := verifyReferencesMap(referencesMap)
		if (isValid == false){
			t.Fatalf("locusMetadataObjectsList contains invalid references map.")
		}
	}

	//TODO: Check to make sure that there are no identical company aliases for different loci

	missingLociList := make([]int64, 0)

	for rsID, _ := range allRSIDsMap{

		_, exists := locusMetadataRSIDsMap[rsID]
		if (exists == false){
			missingLociList = append(missingLociList, rsID)
		}
	}

	if (len(missingLociList) != 0){

		missingLociStringsList := make([]string, 0, len(missingLociList))

		for _, rsID := range missingLociList{

			rsIDString := helpers.ConvertInt64ToString(rsID)

			missingLociStringsList = append(missingLociStringsList, rsIDString)
		}

		missingLociListFormatted := strings.Join(missingLociStringsList, ", ")

		t.Fatalf("locusMetadata is missing loci: " + missingLociListFormatted)
	}
}


/*
// We use this to determine the greatest possible number of variants tested
// This needs to be updated in profileFormat whenever a new monogenic disease is added which exceeds this value
func TestGetHighestPossibleMonogenicDiseaseVariantCount(t *testing.T){

	monogenicDiseases.InitializeMonogenicDiseaseVariables()

	monogenicDiseasesObjectsList, err := monogenicDiseases.GetMonogenicDiseaseObjectsList()
	if (err != nil){
		t.Fatalf("Failed to get monogenic disease objects list: " + err.Error())
	}	

	highestCount := 0

	for _, diseaseObject := range monogenicDiseasesObjectsList{

		diseaseVariantsList := diseaseObject.VariantsList

		diseaseNumberOfVariants := len(diseaseVariantsList)

		if (diseaseNumberOfVariants > highestCount){
			highestCount = diseaseNumberOfVariants
		}
	}

	highestVariantCountString := helpers.ConvertIntToString(highestCount)

	log.Println("Most monogenic disease variants: " + highestVariantCountString)
}

*/