コード例 #1
0
 def test_bool_filtered_query(self):
     ''' Test building and running a filtered boolean query. '''
     query_bool = BoolQuery(must_not_arr=[Query.term("seqid", 2)],
                            should_arr=[RangeQuery("start", gte=10050)])
     query_bool.must([Query.term("id", "rs768019142")]) \
               .should(RangeQuery("start", gte=10054))
     query = ElasticQuery.filtered_bool(Query.match_all(), query_bool, sources=["id", "seqid"])
     elastic = Search(query, idx=ElasticSettings.idx('DEFAULT'))
     self.assertTrue(elastic.search().hits_total == 1, "Elastic filtered query retrieved marker (rs768019142)")
コード例 #2
0
ファイル: criteria.py プロジェクト: D-I-L/django-criteria
    def get_elastic_query(cls, section=None, config=None):
        ''' function to build the elastic query object
        @type  section: string
        @keyword section: The section in the criteria.ini file
        @type  config:  string
        @keyword config: The config object initialized from criteria.ini.
        @return: L{Query}
        '''
        section_config = config[section]
        source_fields = []

        if 'source_fields' in section_config:
            source_fields_str = section_config['source_fields']
            source_fields = source_fields_str.split(',')

        if 'mhc' in section:
            seqid = '6'
            start_range = 25000000
            end_range = 35000000

            seqid_param = section_config['seqid_param']
            start_param = section_config['start_param']
            end_param = section_config['end_param']

        if section == 'is_gene_in_mhc':
            # for region you should make a different query
            # Defined MHC region as chr6:25,000,000..35,000,000

            query = ElasticUtils.range_overlap_query(seqid, start_range, end_range,
                                                     source_fields,
                                                     seqid_param,
                                                     start_param,
                                                     end_param)
        elif section == 'is_marker_in_mhc':
            query_bool = BoolQuery()
            query_bool.must(RangeQuery("start", lte=end_range)) \
                      .must(RangeQuery("start", gte=start_range)) \
                      .must(Query.term("seqid", seqid))
            query = ElasticQuery.filtered_bool(Query.match_all(), query_bool, sources=["id", "seqid", "start"])
        elif section == 'is_region_in_mhc':
            query = ElasticQuery(Query.term("region_name", "MHC"))
        elif section == 'marker_is_gwas_significant_in_ic':
            # build a range query
            gw_sig_p = 0.00000005
            query = ElasticQuery(RangeQuery("p_value", lte=gw_sig_p))
        else:
            if len(source_fields) > 0:
                query = ElasticQuery(Query.match_all(), sources=source_fields)
            else:
                # query = ElasticQuery(Query.match_all())
                return None

        return query
コード例 #3
0
 def test_bool_query(self):
     ''' Test a bool query. '''
     query_bool = BoolQuery()
     highlight = Highlight(["id", "seqid"])
     query_bool.must(Query.term("id", "rs768019142")) \
               .must(RangeQuery("start", gt=1000)) \
               .must_not(Query.match("seqid", "2")) \
               .should(Query.match("seqid", "3")) \
               .should(Query.match("seqid", "1"))
     query = ElasticQuery.bool(query_bool, highlight=highlight)
     elastic = Search(query, idx=ElasticSettings.idx('DEFAULT'))
     self.assertTrue(len(elastic.search().docs) == 1, "Elastic string query retrieved marker (rs768019142)")
コード例 #4
0
    def test_bool_nested_filter(self):
        ''' Test combined Bool filter '''
        query_bool_nest = BoolQuery()
        query_bool_nest.must(Query.match("id", "rs768019142").query_wrap()) \
                       .must(Query.term("seqid", 1))

        query_bool = BoolQuery()
        query_bool.should(query_bool_nest) \
                  .should(Query.term("seqid", 2))
        query = ElasticQuery.filtered_bool(Query.match_all(), query_bool, sources=["id", "seqid", "start"])
        elastic = Search(query, idx=ElasticSettings.idx('DEFAULT'))
        self.assertTrue(elastic.search().hits_total >= 1, "Nested bool filter query")
コード例 #5
0
ファイル: views.py プロジェクト: tcarver/django-chicp
def _build_exon_query(chrom, segmin, segmax, genes):
    # get exonic structure for genes in this section
    geneExons = dict()
    query_bool = BoolQuery()
    query_bool.must([Query.term("seqid", chrom)])
    if len(genes) > 0:
        for g in genes:
            query = ElasticQuery.filtered_bool(Query.query_string(g["gene_id"], fields=["name"]),
                                               query_bool, sources=utils.snpFields)
            elastic = Search(query, idx=getattr(chicp_settings, 'CP_GENE_IDX')+'/exons/', search_from=0, size=2000)
            result = elastic.get_result()
            exons = result['data']
            exons = utils.makeRelative(int(segmin), int(segmax), ['start', 'end'], exons)
            geneExons[g["gene_id"]] = sorted(exons, key=operator.itemgetter("start"))
    return geneExons
コード例 #6
0
ファイル: criteria.py プロジェクト: D-I-L/django-criteria
    def fetch_overlapping_features(cls, build, seqid, start, end, idx=None, idx_type=None, disease_id=None):
        ''' function to create fetch overlapping features for a given stretch of region
            the build info is stored as nested document..so nested query is build
        @type  build: string
        @param build: build info eg: 'GRCh38'
        @type  seqid: string
        @param seqid: chromosome number
        @type  start:  string
        @param start: region start
        @type  end:  string
        @param end: region end
        @type  idx: string
        @param idx: name of the index
        @type  idx_type: string
        @param idx_type: name of the idx type, each criteria is an index type
        @type  disease_id:  string
        @param disease_id: disease code
        '''
        nbuild = build
        start_range = start
        end_range = end

        bool_range = BoolQuery()
        bool_range.must(RangeQuery("build_info.start", lte=start_range)) \
                  .must(RangeQuery("build_info.end", gte=end_range))

        or_filter = OrFilter(RangeQuery("build_info.start", gte=start_range, lte=end_range))

        or_filter.extend(RangeQuery("build_info.end", gte=start_range, lte=end_range)) \
                 .extend(bool_range)

        bool_query = BoolQuery()

        if disease_id:
            qnested_buildinfo = Query.nested('build_info', bool_query)
            bool_query = BoolQuery()
            bool_query.must(Query.term("disease", disease_id.lower())).must(qnested_buildinfo)
            qnested = ElasticQuery(bool_query, sources=['build_info.*',
                                                        'disease_locus',
                                                        'disease',
                                                        'chr_band',
                                                        'species'])

        else:
            bool_query.must(Query.term("build_info.build", nbuild)) \
                  .must(Query.term("build_info.seqid", seqid)) \
                  .filter(or_filter)

            qnested = ElasticQuery(Query.nested('build_info', bool_query), sources=['build_info.*',
                                                                                    'disease_locus',
                                                                                    'disease',
                                                                                    'chr_band',
                                                                                    'species'])

        elastic = Search(qnested, idx=idx, idx_type=idx_type)
        res = elastic.search()
        return res.docs
コード例 #7
0
ファイル: views.py プロジェクト: D-I-L/django-chicp
def chicpeaSearch(request, url):
    queryDict = request.GET
    user = request.user
    targetIdx = queryDict.get("targetIdx")
    blueprint = {}
    hic = []
    addList = []
    searchType = 'gene'
    searchTerm = queryDict.get("searchTerm").upper()
    searchTerm = searchTerm.replace(",", "")
    searchTerm = searchTerm.replace("..", "-")
    searchTerm = searchTerm.replace(" ", "") # Chris suggestion to prevent issue with spaces in queries
    snpTrack = queryDict.get("snp_track")

    (idx_keys_auth, idx_type_keys_auth) = get_authenticated_idx_and_idx_types(
                                            user=user, idx_keys=None, idx_type_keys=None)

    if snpTrack:
        mo = re.match(r"(.*)-(.*)", snpTrack)
        (group, track) = mo.group(1, 2)  # @UnusedVariable
        if group != 'ud' and 'CP_STATS_'+group.upper()+'.'+snpTrack.upper() not in idx_type_keys_auth:
            snpTrack = None

    if targetIdx not in utils.tissues:
        for target in getattr(chicp_settings, 'CP_TARGET'):
            if 'CP_TARGET_'+target not in idx_keys_auth:
                if targetIdx == target:
                    retJSON = {'error': 'Sorry, you do not have permission to view this dataset.'}
                    return JsonResponse(retJSON)
                continue
            elasticJSON = Search(idx=ElasticSettings.idx('CP_TARGET_'+target)).get_mapping(mapping_type="gene_target")
            tissueList = list(elasticJSON[ElasticSettings.idx('CP_TARGET_'+target)]
                              ['mappings']['gene_target']['_meta']['tissue_type'].keys())
            utils.tissues['CP_TARGET_'+target] = tissueList

    if queryDict.get("region") or re.match(r"(.*):(\d+)-(\d+)", searchTerm):
        searchType = 'region'
        region = searchTerm
        if queryDict.get("region"):
            region = queryDict.get("region")
        else:
            searchTerm = ""
        mo = re.match(r"(.*):(\d+)-(\d+)", region)
        (chrom, segmin, segmax) = mo.group(1, 2, 3)
        chrom = chrom.replace('chr', "")
        chrom = chrom.replace('CHR', "")
    if re.search("^rs[0-9]+", searchTerm.lower()):
        searchTerm = searchTerm.lower()
        addList.append(_find_snp_position(snpTrack, searchTerm))
        if addList[0].get("error"):
            return JsonResponse({'error': addList[0]['error']})
        position = addList[0]['end']
        if searchType != 'region':
            searchType = 'snp'

    logger.warn("### "+searchType+" - "+searchTerm+' ###')

    if searchType == 'region':
        query_bool = BoolQuery()
        filter_bool = BoolQuery()
        if searchTerm and len(addList) == 0 and re.match(r"(.*):(\d+)-(\d+)",
                                                         queryDict.get("searchTerm").replace(",", "")) == None:
            query_bool.must([Query.query_string(searchTerm, fields=["name", "ensg"]),
                             Query.term("baitChr", chrom),
                             Query.term("oeChr", chrom),
                             RangeQuery("dist", gte=-2e6, lte=2e6)])
        else:
            query_bool.must([Query.term("baitChr", chrom),
                             Query.term("oeChr", chrom),
                             RangeQuery("dist", gte=-2e6, lte=2e6)])

        query_bool = _add_tissue_filter(query_bool, targetIdx)

        if len(addList) > 0:
            filter_bool.should([BoolQuery(must_arr=[RangeQuery("baitStart", lte=position),
                                                    RangeQuery("baitEnd", gte=position)]),
                                BoolQuery(must_arr=[RangeQuery("oeStart", lte=position),
                                                    RangeQuery("oeEnd", gte=position)])])
        else:
            filter_bool.should([BoolQuery(must_arr=[RangeQuery("baitStart", gte=segmin, lte=segmax),
                                                    RangeQuery("baitEnd", gte=segmin, lte=segmax)]),
                                BoolQuery(must_arr=[RangeQuery("oeStart", gte=segmin, lte=segmax),
                                                    RangeQuery("oeEnd", gte=segmin, lte=segmax)])])

        query = ElasticQuery.filtered_bool(query_bool, filter_bool,
                                           sources=utils.hicFields + utils.tissues['CP_TARGET_'+targetIdx])
        (hic, v1, v2) = _build_hic_query(query, targetIdx, segmin, segmax)  # @UnusedVariable

        if "error" in hic:
            return JsonResponse(hic)
        if len(hic) == 0:
            retJSON = {'error': queryDict.get("searchTerm")+' does not overlap any bait/target regions in this dataset.'}
            return JsonResponse(retJSON)

    elif searchType == 'snp':
        if len(addList) > 0:
            chrom = addList[0]['chr']

            query_bool = BoolQuery()
            query_bool.must([Query.term("baitChr", chrom),
                             Query.term("oeChr", chrom),
                             RangeQuery("dist", gte=-2e6, lte=2e6)])
            query_bool = _add_tissue_filter(query_bool, targetIdx)

            filter_bool = BoolQuery()
            filter_bool.should([BoolQuery(must_arr=[RangeQuery("baitStart", lte=position),
                                                    RangeQuery("baitEnd", gte=position)]),
                                BoolQuery(must_arr=[RangeQuery("oeStart", lte=position),
                                                    RangeQuery("oeEnd", gte=position)])])

            query = ElasticQuery.filtered_bool(query_bool, filter_bool,
                                               sources=utils.hicFields + utils.tissues['CP_TARGET_'+targetIdx])
            hic, segmin, segmax = _build_hic_query(query, targetIdx)

            if "error" in hic:
                return JsonResponse(hic)
            if len(hic) == 0:
                retJSON = {'error': 'Marker '+searchTerm+' does not overlap any bait/target regions in this dataset.'}
                return JsonResponse(retJSON)
    else:
        # geneQuery = ElasticQuery.query_string(searchTerm, fields=["gene_name"])
        geneQuery = ElasticQuery.filtered(Query.match_all(), Filter(Query.match("gene_name", searchTerm).query_wrap()))
        resultObj = Search(idx=getattr(chicp_settings, 'CP_GENE_IDX') + '/genes/',
                           search_query=geneQuery, size=0, qsort=Sort('seqid:asc,start')).search()
        if resultObj.hits_total > 1:
            geneResults = []
            resultObj2 = Search(idx=getattr(chicp_settings, 'CP_GENE_IDX') + '/genes/', search_query=geneQuery,
                                size=(resultObj.hits_total+1), qsort=Sort('seqid:asc,start')).search()

            docs = resultObj2.docs
            gene_ids = [getattr(doc, 'attr')['gene_id'][1:-1] for doc in docs]

            query = ElasticQuery.filtered(Query.match_all(), TermsFilter.get_terms_filter('ensg', gene_ids))
            agg = Agg('ensg_agg', "terms", {"field": "ensg", "size": 0})
            res = Search(idx=ElasticSettings.idx('CP_TARGET_'+targetIdx), search_query=query, aggs=Aggs(agg),
                         size=0).search()

            ensg_count = res.aggs['ensg_agg'].get_buckets()
            gene_ids = [g['key'] for g in ensg_count]

            for d in resultObj2.docs:
                if getattr(d, "attr")["gene_id"].replace('\"', '') in gene_ids:
                    geneResults.append({
                        'gene_name': getattr(d, "attr")["gene_name"].replace('\"', ''),
                        'gene_id': getattr(d, "attr")["gene_id"].replace('\"', ''),
                        'location': "chr" + getattr(d, "seqid") + ":" +
                        locale.format_string("%d", getattr(d, "start"), grouping=True) + ".." +
                        locale.format_string("%d", getattr(d, "end"), grouping=True),
                    })

            if len(geneResults) == 0:
                retJSON = {'error': 'Gene name '+searchTerm+' not found in this dataset.'}
                return JsonResponse(retJSON)
            elif len(geneResults) > 1:
                retJSON = {
                    'error': 'Gene name <strong>'+searchTerm+'</strong> returns too many hits, please select your prefered result from the list below.',
                    'results': geneResults,
                    'cols': ['HGNC Symbol', 'Ensembl Gene ID', 'Location']
                }
                return JsonResponse(retJSON)

        query_bool = BoolQuery()
        query_bool.must([RangeQuery("dist", gte=-2e6, lte=2e6)])
        query_bool = _add_tissue_filter(query_bool, targetIdx)
        query = ElasticQuery.filtered_bool(Query.query_string(searchTerm, fields=["name", "ensg", "oeName"]),
                                           query_bool, sources=utils.hicFields + utils.tissues['CP_TARGET_'+targetIdx])

        (hic, segmin, segmax) = _build_hic_query(query, targetIdx)

        if "error" in hic:
            return JsonResponse(hic)
        if len(hic) == 0:
            retJSON = {'error': 'Gene name '+searchTerm+' not found in this dataset.'}
            return JsonResponse(retJSON)
        chrom = hic[0]['baitChr']

    try:
        chrom
    except NameError:
        retJSON = {'error': 'No chromosome defined for search'}
        return JsonResponse(retJSON)

    # get genes based on this segment
    genes = _build_gene_query(chrom, segmin, segmax)
    (snps, snpMeta) = _build_snp_query(snpTrack, chrom, segmin, segmax)
    frags = _build_frags_query(getattr(chicp_settings, 'DEFAULT_FRAG'), chrom, segmin, segmax)

    addList = utils.makeRelative(int(segmin), int(segmax), ['start', 'end'], addList)

    retJSON = {"hic": hic,
               "frags": frags,
               "meta": {"ostart": int(segmin),
                        "oend": int(segmax),
                        "rstart": 1,
                        "rend": int(segmax) - int(segmin),
                        "rchr": str(chrom),
                        "tissues": utils.tissues['CP_TARGET_'+targetIdx]},
               "snps": snps,
               "snp_meta": snpMeta,
               "genes": genes,
               "region": str(chrom) + ":" + str(segmin) + "-" + str(segmax),
               "blueprint": blueprint,
               "extra": addList
               }

    response = JsonResponse(retJSON)
    return response
コード例 #8
0
ファイル: views.py プロジェクト: tcarver/django-chicp
def chicpeaSearch(request, url):
    queryDict = request.GET
    targetIdx = queryDict.get("targetIdx")
    blueprint = {}
    hic = []
    addList = []
    searchType = 'gene'
    searchTerm = queryDict.get("searchTerm").upper()

    if targetIdx not in utils.tissues:
        for idx in getattr(chicp_settings, 'TARGET_IDXS'):
            elasticJSON = Search(idx=idx).get_mapping(mapping_type="gene_target")
            tissueList = list(elasticJSON[idx]['mappings']['gene_target']['_meta']['tissue_type'].keys())
            utils.tissues[idx] = tissueList

    if queryDict.get("region") or re.match(r"(.*):(\d+)-(\d+)", queryDict.get("searchTerm")):
        searchType = 'region'
        region = queryDict.get("searchTerm")
        if queryDict.get("region"):
            region = queryDict.get("region")
        else:
            searchTerm = ""
        mo = re.match(r"(.*):(\d+)-(\d+)", region)
        (chrom, segmin, segmax) = mo.group(1, 2, 3)
        chrom = chrom.replace('chr', "")
    if re.search("^rs[0-9]+", queryDict.get("searchTerm").lower()):
        searchTerm = queryDict.get("searchTerm").lower()
        addList.append(_find_snp_position(queryDict.get("snp_track"), searchTerm))
        if addList[0].get("error"):
            return JsonResponse({'error': addList[0]['error']})
        position = addList[0]['end']
        if searchType != 'region':
            searchType = 'snp'

    logger.warn("### "+searchType+" - "+searchTerm+' ###')

    if searchType == 'region':
        query_bool = BoolQuery()
        filter_bool = BoolQuery()
        if searchTerm and len(addList) == 0 and re.match(r"(.*):(\d+)-(\d+)", queryDict.get("searchTerm")) == None:
            query_bool.must([Query.query_string(searchTerm, fields=["name", "ensg"]),
                             Query.term("baitChr", chrom),
                             Query.term("oeChr", chrom),
                             RangeQuery("dist", gte=-2e6, lte=2e6)])
        else:
            query_bool.must([Query.term("baitChr", chrom),
                             Query.term("oeChr", chrom),
                             RangeQuery("dist", gte=-2e6, lte=2e6)])

        query_bool = _add_tissue_filter(query_bool, targetIdx)

        if len(addList) > 0:
            filter_bool.should([BoolQuery(must_arr=[RangeQuery("baitStart", lte=position),
                                                    RangeQuery("baitEnd", gte=position)]),
                                BoolQuery(must_arr=[RangeQuery("oeStart", lte=position),
                                                    RangeQuery("oeEnd", gte=position)])])
        else:
            filter_bool.should([BoolQuery(must_arr=[RangeQuery("baitStart", gte=segmin, lte=segmax),
                                                    RangeQuery("baitEnd", gte=segmin, lte=segmax)]),
                                BoolQuery(must_arr=[RangeQuery("oeStart", gte=segmin, lte=segmax),
                                                    RangeQuery("oeEnd", gte=segmin, lte=segmax)])])

        query = ElasticQuery.filtered_bool(query_bool, filter_bool, sources=utils.hicFields + utils.tissues[targetIdx])
        (hic, v1, v2) = _build_hic_query(query, targetIdx, segmin, segmax)
        # print(hic)

        if len(hic) == 0:
            retJSON = {'error': queryDict.get("searchTerm")+' does not overlap any bait/target regions in this dataset.'}
            return JsonResponse(retJSON)

    elif searchType == 'snp':
        if len(addList) > 0:
            chrom = addList[0]['chr']

            query_bool = BoolQuery()
            query_bool.must([Query.term("baitChr", chrom),
                             Query.term("oeChr", chrom),
                             RangeQuery("dist", gte=-2e6, lte=2e6)])
            query_bool = _add_tissue_filter(query_bool, targetIdx)

            filter_bool = BoolQuery()
            filter_bool.should([BoolQuery(must_arr=[RangeQuery("baitStart", lte=position),
                                                    RangeQuery("baitEnd", gte=position)]),
                                BoolQuery(must_arr=[RangeQuery("oeStart", lte=position),
                                                    RangeQuery("oeEnd", gte=position)])])

            query = ElasticQuery.filtered_bool(query_bool, filter_bool,
                                               sources=utils.hicFields + utils.tissues[targetIdx])
            hic, segmin, segmax = _build_hic_query(query, targetIdx)

            if len(hic) == 0:
                retJSON = {'error': 'Marker '+searchTerm+' does not overlap any bait/target regions in this dataset.'}
                return JsonResponse(retJSON)
    else:
        query_bool = BoolQuery()
        query_bool.must([RangeQuery("dist", gte=-2e6, lte=2e6)])
        query_bool = _add_tissue_filter(query_bool, targetIdx)
        query = ElasticQuery.filtered_bool(Query.query_string(searchTerm, fields=["name", "ensg", "oeName"]),
                                           query_bool, sources=utils.hicFields + utils.tissues[targetIdx])

        hic, segmin, segmax = _build_hic_query(query, targetIdx)

        if len(hic) == 0:
            retJSON = {'error': 'Gene name '+searchTerm+' not found in this dataset.'}
            return JsonResponse(retJSON)
        chrom = hic[0]['baitChr']

    try:
        chrom
    except NameError:
        retJSON = {'error': 'No chromosome defined for search'}
        return JsonResponse(retJSON)

    # get genes based on this segment
    genes = _build_gene_query(chrom, segmin, segmax)
    (snps, snpMeta) = _build_snp_query(queryDict.get("snp_track"), chrom, segmin, segmax)
    frags = _build_frags_query(getattr(chicp_settings, 'DEFAULT_FRAG'), chrom, segmin, segmax)

    addList = utils.makeRelative(int(segmin), int(segmax), ['start', 'end'], addList)

    retJSON = {"hic": hic,
               "frags": frags,
               "meta": {"ostart": int(segmin),
                        "oend": int(segmax),
                        "rstart": 1,
                        "rend": int(segmax) - int(segmin),
                        "rchr": str(chrom),
                        "tissues": utils.tissues[targetIdx]},
               "snps": snps,
               "snp_meta": snpMeta,
               "genes": genes,
               "region": str(chrom) + ":" + str(segmin) + "-" + str(segmax),
               "blueprint": blueprint,
               "extra": addList
               }

    response = JsonResponse(retJSON)
    return response
コード例 #9
0
def chicpeaSearch(request, url):
    queryDict = request.GET
    user = request.user
    targetIdx = queryDict.get("targetIdx")
    blueprint = {}
    hic = []
    addList = []
    searchType = 'gene'
    searchTerm = queryDict.get("searchTerm").upper()
    searchTerm = searchTerm.replace(",", "")
    searchTerm = searchTerm.replace("..", "-")
    snpTrack = queryDict.get("snp_track")

    (idx_keys_auth, idx_type_keys_auth) = get_authenticated_idx_and_idx_types(
                                            user=user, idx_keys=None, idx_type_keys=None)

    if snpTrack:
        mo = re.match(r"(.*)-(.*)", snpTrack)
        (group, track) = mo.group(1, 2)  # @UnusedVariable
        if group != 'ud' and 'CP_STATS_'+group.upper()+'.'+snpTrack.upper() not in idx_type_keys_auth:
            snpTrack = None

    if targetIdx not in utils.tissues:
        for target in getattr(chicp_settings, 'CP_TARGET'):
            if 'CP_TARGET_'+target not in idx_keys_auth:
                if targetIdx == target:
                    retJSON = {'error': 'Sorry, you do not have permission to view this dataset.'}
                    return JsonResponse(retJSON)
                continue
            elasticJSON = Search(idx=ElasticSettings.idx('CP_TARGET_'+target)).get_mapping(mapping_type="gene_target")
            tissueList = list(elasticJSON[ElasticSettings.idx('CP_TARGET_'+target)]
                              ['mappings']['gene_target']['_meta']['tissue_type'].keys())
            utils.tissues['CP_TARGET_'+target] = tissueList

    if queryDict.get("region") or re.match(r"(.*):(\d+)-(\d+)", searchTerm):
        searchType = 'region'
        region = searchTerm
        if queryDict.get("region"):
            region = queryDict.get("region")
        else:
            searchTerm = ""
        mo = re.match(r"(.*):(\d+)-(\d+)", region)
        (chrom, segmin, segmax) = mo.group(1, 2, 3)
        chrom = chrom.replace('chr', "")
        chrom = chrom.replace('CHR', "")
    if re.search("^rs[0-9]+", searchTerm.lower()):
        searchTerm = searchTerm.lower()
        addList.append(_find_snp_position(snpTrack, searchTerm))
        if addList[0].get("error"):
            return JsonResponse({'error': addList[0]['error']})
        position = addList[0]['end']
        if searchType != 'region':
            searchType = 'snp'

    logger.warn("### "+searchType+" - "+searchTerm+' ###')

    if searchType == 'region':
        query_bool = BoolQuery()
        filter_bool = BoolQuery()
        if searchTerm and len(addList) == 0 and re.match(r"(.*):(\d+)-(\d+)",
                                                         queryDict.get("searchTerm").replace(",", "")) == None:
            query_bool.must([Query.query_string(searchTerm, fields=["name", "ensg"]),
                             Query.term("baitChr", chrom),
                             Query.term("oeChr", chrom),
                             RangeQuery("dist", gte=-2e6, lte=2e6)])
        else:
            query_bool.must([Query.term("baitChr", chrom),
                             Query.term("oeChr", chrom),
                             RangeQuery("dist", gte=-2e6, lte=2e6)])

        query_bool = _add_tissue_filter(query_bool, targetIdx)

        if len(addList) > 0:
            filter_bool.should([BoolQuery(must_arr=[RangeQuery("baitStart", lte=position),
                                                    RangeQuery("baitEnd", gte=position)]),
                                BoolQuery(must_arr=[RangeQuery("oeStart", lte=position),
                                                    RangeQuery("oeEnd", gte=position)])])
        else:
            filter_bool.should([BoolQuery(must_arr=[RangeQuery("baitStart", gte=segmin, lte=segmax),
                                                    RangeQuery("baitEnd", gte=segmin, lte=segmax)]),
                                BoolQuery(must_arr=[RangeQuery("oeStart", gte=segmin, lte=segmax),
                                                    RangeQuery("oeEnd", gte=segmin, lte=segmax)])])

        query = ElasticQuery.filtered_bool(query_bool, filter_bool,
                                           sources=utils.hicFields + utils.tissues['CP_TARGET_'+targetIdx])
        (hic, v1, v2) = _build_hic_query(query, targetIdx, segmin, segmax)  # @UnusedVariable

        if "error" in hic:
            return JsonResponse(hic)
        if len(hic) == 0:
            retJSON = {'error': queryDict.get("searchTerm")+' does not overlap any bait/target regions in this dataset.'}
            return JsonResponse(retJSON)

    elif searchType == 'snp':
        if len(addList) > 0:
            chrom = addList[0]['chr']

            query_bool = BoolQuery()
            query_bool.must([Query.term("baitChr", chrom),
                             Query.term("oeChr", chrom),
                             RangeQuery("dist", gte=-2e6, lte=2e6)])
            query_bool = _add_tissue_filter(query_bool, targetIdx)

            filter_bool = BoolQuery()
            filter_bool.should([BoolQuery(must_arr=[RangeQuery("baitStart", lte=position),
                                                    RangeQuery("baitEnd", gte=position)]),
                                BoolQuery(must_arr=[RangeQuery("oeStart", lte=position),
                                                    RangeQuery("oeEnd", gte=position)])])

            query = ElasticQuery.filtered_bool(query_bool, filter_bool,
                                               sources=utils.hicFields + utils.tissues['CP_TARGET_'+targetIdx])
            hic, segmin, segmax = _build_hic_query(query, targetIdx)

            if "error" in hic:
                return JsonResponse(hic)
            if len(hic) == 0:
                retJSON = {'error': 'Marker '+searchTerm+' does not overlap any bait/target regions in this dataset.'}
                return JsonResponse(retJSON)
    else:
        # geneQuery = ElasticQuery.query_string(searchTerm, fields=["gene_name"])
        geneQuery = ElasticQuery.filtered(Query.match_all(), Filter(Query.match("gene_name", searchTerm).query_wrap()))
        resultObj = Search(idx=getattr(chicp_settings, 'CP_GENE_IDX') + '/genes/',
                           search_query=geneQuery, size=0, qsort=Sort('seqid:asc,start')).search()
        if resultObj.hits_total > 1:
            geneResults = []
            resultObj2 = Search(idx=getattr(chicp_settings, 'CP_GENE_IDX') + '/genes/', search_query=geneQuery,
                                size=(resultObj.hits_total+1), qsort=Sort('seqid:asc,start')).search()

            docs = resultObj2.docs
            gene_ids = [getattr(doc, 'attr')['gene_id'][1:-1] for doc in docs]

            query = ElasticQuery.filtered(Query.match_all(), TermsFilter.get_terms_filter('ensg', gene_ids))
            agg = Agg('ensg_agg', "terms", {"field": "ensg", "size": 0})
            res = Search(idx=ElasticSettings.idx('CP_TARGET_'+targetIdx), search_query=query, aggs=Aggs(agg),
                         size=0).search()

            ensg_count = res.aggs['ensg_agg'].get_buckets()
            gene_ids = [g['key'] for g in ensg_count]

            for d in resultObj2.docs:
                if getattr(d, "attr")["gene_id"].replace('\"', '') in gene_ids:
                    geneResults.append({
                        'gene_name': getattr(d, "attr")["gene_name"].replace('\"', ''),
                        'gene_id': getattr(d, "attr")["gene_id"].replace('\"', ''),
                        'location': "chr" + getattr(d, "seqid") + ":" +
                        locale.format_string("%d", getattr(d, "start"), grouping=True) + ".." +
                        locale.format_string("%d", getattr(d, "end"), grouping=True),
                    })

            if len(geneResults) == 0:
                retJSON = {'error': 'Gene name '+searchTerm+' not found in this dataset.'}
                return JsonResponse(retJSON)
            elif len(geneResults) > 1:
                retJSON = {
                    'error': 'Gene name <strong>'+searchTerm+'</strong> returns too many hits, please select your prefered result from the list below.',
                    'results': geneResults,
                    'cols': ['HGNC Symbol', 'Ensembl Gene ID', 'Location']
                }
                return JsonResponse(retJSON)

        query_bool = BoolQuery()
        query_bool.must([RangeQuery("dist", gte=-2e6, lte=2e6)])
        query_bool = _add_tissue_filter(query_bool, targetIdx)
        query = ElasticQuery.filtered_bool(Query.query_string(searchTerm, fields=["name", "ensg", "oeName"]),
                                           query_bool, sources=utils.hicFields + utils.tissues['CP_TARGET_'+targetIdx])

        (hic, segmin, segmax) = _build_hic_query(query, targetIdx)

        if "error" in hic:
            return JsonResponse(hic)
        if len(hic) == 0:
            retJSON = {'error': 'Gene name '+searchTerm+' not found in this dataset.'}
            return JsonResponse(retJSON)
        chrom = hic[0]['baitChr']

    try:
        chrom
    except NameError:
        retJSON = {'error': 'No chromosome defined for search'}
        return JsonResponse(retJSON)

    # get genes based on this segment
    genes = _build_gene_query(chrom, segmin, segmax)
    (snps, snpMeta) = _build_snp_query(snpTrack, chrom, segmin, segmax)
    frags = _build_frags_query(getattr(chicp_settings, 'DEFAULT_FRAG'), chrom, segmin, segmax)

    addList = utils.makeRelative(int(segmin), int(segmax), ['start', 'end'], addList)

    retJSON = {"hic": hic,
               "frags": frags,
               "meta": {"ostart": int(segmin),
                        "oend": int(segmax),
                        "rstart": 1,
                        "rend": int(segmax) - int(segmin),
                        "rchr": str(chrom),
                        "tissues": utils.tissues['CP_TARGET_'+targetIdx]},
               "snps": snps,
               "snp_meta": snpMeta,
               "genes": genes,
               "region": str(chrom) + ":" + str(segmin) + "-" + str(segmax),
               "blueprint": blueprint,
               "extra": addList
               }

    response = JsonResponse(retJSON)
    return response