import json import math import collections import requests from redis import Redis from flask import g from gn3.utility.db_tools import escape from gn3.utility.db_tools import mescape from gn3.utility.db_tools import create_in_clause from gn3.utility.tools import locate_ignore_error from gn3.db.calls import fetch1 from gn3.db.calls import fetchone from gn3.db.webqtlDatabaseFunction import retrieve_species from gn3.utility import chunks from gn3.utility import get_group_samplelists from gn3.base.species import TheSpecies r = Redis() # should probably move this to its own configuration files USE_REDIS = True # todo move to config file GN2_BASE_URL = "https://genenetwork.org/" DS_NAME_MAP = {} # pylint: disable-all #todo file not linted # pylint: disable=C0103 def create_dataset(dataset_name, dataset_type=None, get_samplelist=True, group_name=None): if dataset_name == "Temp": dataset_type = "Temp" if dataset_type is None: dataset_type = Dataset_Getter(dataset_name) dataset_ob = DS_NAME_MAP[dataset_type] dataset_class = globals()[dataset_ob] if dataset_type == "Temp": results = dataset_class(dataset_name, get_samplelist, group_name) else: results = dataset_class(dataset_name, get_samplelist) return results class DatasetType: def __init__(self, redis_instance): self.redis_instance = redis_instance self.datasets = {} data = self.redis_instance.get("dataset_structure") if data: self.datasets = json.loads(data) else: try: data = json.loads(requests.get( GN2_BASE_URL + "/api/v_pre1/gen_dropdown", timeout=5).content) # todo:Refactor code below n^4 loop for species in data["datasets"]: for group in data["datasets"][species]: for dataset_type in data['datasets'][species][group]: for dataset in data['datasets'][species][group][dataset_type]: short_dataset_name = dataset[1] if dataset_type == "Phenotypes": new_type = "Publish" elif dataset_type == "Genotypes": new_type = "Geno" else: new_type = "ProbeSet" self.datasets[short_dataset_name] = new_type except Exception as e: raise e self.redis_instance.set( "dataset_structure", json.dumps(self.datasets)) def set_dataset_key(self, t, name): """If name is not in the object's dataset dictionary, set it, and update dataset_structure in Redis args: t: Type of dataset structure which can be: 'mrna_expr', 'pheno', 'other_pheno', 'geno' name: The name of the key to inserted in the datasets dictionary """ sql_query_mapping = { 'mrna_expr': ("""SELECT ProbeSetFreeze.Id FROM """ + """ProbeSetFreeze WHERE ProbeSetFreeze.Name = "{}" """), 'pheno': ("""SELECT InfoFiles.GN_AccesionId """ + """FROM InfoFiles, PublishFreeze, InbredSet """ + """WHERE InbredSet.Name = '{}' AND """ + """PublishFreeze.InbredSetId = InbredSet.Id AND """ + """InfoFiles.InfoPageName = PublishFreeze.Name"""), 'other_pheno': ("""SELECT PublishFreeze.Name """ + """FROM PublishFreeze, InbredSet """ + """WHERE InbredSet.Name = '{}' AND """ + """PublishFreeze.InbredSetId = InbredSet.Id"""), 'geno': ("""SELECT GenoFreeze.Id FROM GenoFreeze WHERE """ + """GenoFreeze.Name = "{}" """) } dataset_name_mapping = { "mrna_expr": "ProbeSet", "pheno": "Publish", "other_pheno": "Publish", "geno": "Geno", } group_name = name if t in ['pheno', 'other_pheno']: group_name = name.replace("Publish", "") results = g.db.execute( sql_query_mapping[t].format(group_name)).fetchone() if results: self.datasets[name] = dataset_name_mapping[t] self.redis_instance.set( "dataset_structure", json.dumps(self.datasets)) return True return None def __call__(self, name): if name not in self.datasets: for t in ["mrna_expr", "pheno", "other_pheno", "geno"]: if(self.set_dataset_key(t, name)): # This has side-effects, with the end result being a truth-y value break return self.datasets.get(name, None) # Do the intensive work at startup one time only # could replace the code below Dataset_Getter = DatasetType(r) class DatasetGroup: """ Each group has multiple datasets; each species has multiple groups. For example, Mouse has multiple groups (BXD, BXA, etc), and each group has multiple datasets associated with it. """ def __init__(self, dataset, name=None): """This sets self.group and self.group_id""" if name == None: self.name, self.id, self.genetic_type = fetchone( dataset.query_for_group) else: self.name, self.id, self.genetic_type = fetchone( "SELECT InbredSet.Name, InbredSet.Id, InbredSet.GeneticType FROM InbredSet where Name='%s'" % name) if self.name == 'BXD300': self.name = "BXD" self.f1list = None self.parlist = None self.get_f1_parent_strains() # remove below not used in correlation self.mapping_id, self.mapping_names = self.get_mapping_methods() self.species = retrieve_species(self.name) def get_f1_parent_strains(self): try: # should import ParInfo raise e # NL, 07/27/2010. ParInfo has been moved from webqtlForm.py to webqtlUtil.py; f1, f12, maternal, paternal = webqtlUtil.ParInfo[self.name] except Exception as e: f1 = f12 = maternal = paternal = None if f1 and f12: self.f1list = [f1, f12] if maternal and paternal: self.parlist = [maternal, paternal] def get_mapping_methods(self): mapping_id = g.db.execute( "select MappingMethodId from InbredSet where Name= '%s'" % self.name).fetchone()[0] if mapping_id == "1": mapping_names = ["GEMMA", "QTLReaper", "R/qtl"] elif mapping_id == "2": mapping_names = ["GEMMA"] elif mapping_id == "3": mapping_names = ["R/qtl"] elif mapping_id == "4": mapping_names = ["GEMMA", "PLINK"] else: mapping_names = [] return mapping_id, mapping_names def get_samplelist(self): result = None key = "samplelist:v3:" + self.name if USE_REDIS: result = r.get(key) if result is not None: self.samplelist = json.loads(result) else: # logger.debug("Cache not hit") # should enable logger genotype_fn = locate_ignore_error(self.name+".geno", 'genotype') if genotype_fn: self.samplelist = get_group_samplelists.get_samplelist( "geno", genotype_fn) else: self.samplelist = None if USE_REDIS: r.set(key, json.dumps(self.samplelist)) r.expire(key, 60*5) class DataSet: """ DataSet class defines a dataset in webqtl, can be either Microarray, Published phenotype, genotype, or user input dataset(temp) """ def __init__(self, name, get_samplelist=True, group_name=None): assert name, "Need a name" self.name = name self.id = None self.shortname = None self.fullname = None self.type = None self.data_scale = None # ZS: For example log2 self.setup() if self.type == "Temp": # Need to supply group name as input if temp trait # sets self.group and self.group_id and gets genotype self.group = DatasetGroup(self, name=group_name) else: self.check_confidentiality() self.retrieve_other_names() # sets self.group and self.group_id and gets genotype self.group = DatasetGroup(self) self.accession_id = self.get_accession_id() if get_samplelist == True: self.group.get_samplelist() self.species = TheSpecies(self) def get_desc(self): """Gets overridden later, at least for Temp...used by trait's get_given_name""" return None def get_accession_id(self): if self.type == "Publish": results = g.db.execute("""select InfoFiles.GN_AccesionId from InfoFiles, PublishFreeze, InbredSet where InbredSet.Name = %s and PublishFreeze.InbredSetId = InbredSet.Id and InfoFiles.InfoPageName = PublishFreeze.Name and PublishFreeze.public > 0 and PublishFreeze.confidentiality < 1 order by PublishFreeze.CreateTime desc""", (self.group.name)).fetchone() elif self.type == "Geno": results = g.db.execute("""select InfoFiles.GN_AccesionId from InfoFiles, GenoFreeze, InbredSet where InbredSet.Name = %s and GenoFreeze.InbredSetId = InbredSet.Id and InfoFiles.InfoPageName = GenoFreeze.ShortName and GenoFreeze.public > 0 and GenoFreeze.confidentiality < 1 order by GenoFreeze.CreateTime desc""", (self.group.name)).fetchone() else: results = None if results != None: return str(results[0]) else: return "None" def retrieve_other_names(self): """This method fetches the the dataset names in search_result. If the data set name parameter is not found in the 'Name' field of the data set table, check if it is actually the FullName or ShortName instead. This is not meant to retrieve the data set info if no name at all is passed. """ try: if self.type == "ProbeSet": query_args = tuple(escape(x) for x in ( self.name, self.name, self.name)) self.id, self.name, self.fullname, self.shortname, self.data_scale, self.tissue = fetch1(""" SELECT ProbeSetFreeze.Id, ProbeSetFreeze.Name, ProbeSetFreeze.FullName, ProbeSetFreeze.ShortName, ProbeSetFreeze.DataScale, Tissue.Name FROM ProbeSetFreeze, ProbeFreeze, Tissue WHERE ProbeSetFreeze.ProbeFreezeId = ProbeFreeze.Id AND ProbeFreeze.TissueId = Tissue.Id AND (ProbeSetFreeze.Name = '%s' OR ProbeSetFreeze.FullName = '%s' OR ProbeSetFreeze.ShortName = '%s') """ % (query_args), "/dataset/"+self.name+".json", lambda r: (r["id"], r["name"], r["full_name"], r["short_name"], r["data_scale"], r["tissue"]) ) else: query_args = tuple(escape(x) for x in ( (self.type + "Freeze"), self.name, self.name, self.name)) self.tissue = "N/A" self.id, self.name, self.fullname, self.shortname = fetchone(""" SELECT Id, Name, FullName, ShortName FROM %s WHERE (Name = '%s' OR FullName = '%s' OR ShortName = '%s') """ % (query_args)) except TypeError as e: logger.debug( "Dataset {} is not yet available in GeneNetwork.".format(self.name)) pass def get_trait_data(self, sample_list=None): if sample_list: self.samplelist = sample_list else: self.samplelist = self.group.samplelist if self.group.parlist != None and self.group.f1list != None: if (self.group.parlist + self.group.f1list) in self.samplelist: self.samplelist += self.group.parlist + self.group.f1list query = """ SELECT Strain.Name, Strain.Id FROM Strain, Species WHERE Strain.Name IN {} and Strain.SpeciesId=Species.Id and Species.name = '{}' """.format(create_in_clause(self.samplelist), *mescape(self.group.species)) # logger.sql(query) results = dict(g.db.execute(query).fetchall()) sample_ids = [results[item] for item in self.samplelist] # MySQL limits the number of tables that can be used in a join to 61, # so we break the sample ids into smaller chunks # Postgres doesn't have that limit, so we can get rid of this after we transition chunk_size = 50 number_chunks = int(math.ceil(len(sample_ids) / chunk_size)) trait_sample_data = [] for sample_ids_step in chunks.divide_into_chunks(sample_ids, number_chunks): if self.type == "Publish": dataset_type = "Phenotype" else: dataset_type = self.type temp = ['T%s.value' % item for item in sample_ids_step] if self.type == "Publish": query = "SELECT {}XRef.Id,".format(escape(self.type)) else: query = "SELECT {}.Name,".format(escape(dataset_type)) data_start_pos = 1 query += ', '.join(temp) query += ' FROM ({}, {}XRef, {}Freeze) '.format(*mescape(dataset_type, self.type, self.type)) for item in sample_ids_step: query += """ left join {}Data as T{} on T{}.Id = {}XRef.DataId and T{}.StrainId={}\n """.format(*mescape(self.type, item, item, self.type, item, item)) if self.type == "Publish": query += """ WHERE {}XRef.InbredSetId = {}Freeze.InbredSetId and {}Freeze.Name = '{}' and {}.Id = {}XRef.{}Id order by {}.Id """.format(*mescape(self.type, self.type, self.type, self.name, dataset_type, self.type, dataset_type, dataset_type)) else: query += """ WHERE {}XRef.{}FreezeId = {}Freeze.Id and {}Freeze.Name = '{}' and {}.Id = {}XRef.{}Id order by {}.Id """.format(*mescape(self.type, self.type, self.type, self.type, self.name, dataset_type, self.type, self.type, dataset_type)) results = g.db.execute(query).fetchall() trait_sample_data.append(results) trait_count = len(trait_sample_data[0]) self.trait_data = collections.defaultdict(list) # put all of the separate data together into a dictionary where the keys are # trait names and values are lists of sample values for trait_counter in range(trait_count): trait_name = trait_sample_data[0][trait_counter][0] for chunk_counter in range(int(number_chunks)): self.trait_data[trait_name] += ( trait_sample_data[chunk_counter][trait_counter][data_start_pos:]) class MrnaAssayDataSet(DataSet): ''' An mRNA Assay is a quantitative assessment (assay) associated with an mRNA trait This used to be called ProbeSet, but that term only refers specifically to the Affymetrix platform and is far too specific. ''' DS_NAME_MAP['ProbeSet'] = 'MrnaAssayDataSet' def setup(self): # Fields in the database table self.search_fields = ['Name', 'Description', 'Probe_Target_Description', 'Symbol', 'Alias', 'GenbankId', 'UniGeneId', 'RefSeq_TranscriptId'] # Find out what display_fields is self.display_fields = ['name', 'symbol', 'description', 'probe_target_description', 'chr', 'mb', 'alias', 'geneid', 'genbankid', 'unigeneid', 'omim', 'refseq_transcriptid', 'blatseq', 'targetseq', 'chipid', 'comments', 'strand_probe', 'strand_gene', 'proteinid', 'uniprotid', 'probe_set_target_region', 'probe_set_specificity', 'probe_set_blat_score', 'probe_set_blat_mb_start', 'probe_set_blat_mb_end', 'probe_set_strand', 'probe_set_note_by_rw', 'flag'] # Fields displayed in the search results table header self.header_fields = ['Index', 'Record', 'Symbol', 'Description', 'Location', 'Mean', 'Max LRS', 'Max LRS Location', 'Additive Effect'] # Todo: Obsolete or rename this field self.type = 'ProbeSet' self.query_for_group = ''' SELECT InbredSet.Name, InbredSet.Id, InbredSet.GeneticType FROM InbredSet, ProbeSetFreeze, ProbeFreeze WHERE ProbeFreeze.InbredSetId = InbredSet.Id AND ProbeFreeze.Id = ProbeSetFreeze.ProbeFreezeId AND ProbeSetFreeze.Name = "%s" ''' % escape(self.name) def check_confidentiality(self): return geno_mrna_confidentiality(self) def get_trait_info(self, trait_list=None, species=''): # Note: setting trait_list to [] is probably not a great idea. if not trait_list: trait_list = [] for this_trait in trait_list: if not this_trait.haveinfo: this_trait.retrieveInfo(QTL=1) if not this_trait.symbol: this_trait.symbol = "N/A" # XZ, 12/08/2008: description # XZ, 06/05/2009: Rob asked to add probe target description description_string = str( str(this_trait.description).strip(codecs.BOM_UTF8), 'utf-8') target_string = str( str(this_trait.probe_target_description).strip(codecs.BOM_UTF8), 'utf-8') if len(description_string) > 1 and description_string != 'None': description_display = description_string else: description_display = this_trait.symbol if (len(description_display) > 1 and description_display != 'N/A' and len(target_string) > 1 and target_string != 'None'): description_display = description_display + '; ' + target_string.strip() # Save it for the jinja2 template this_trait.description_display = description_display if this_trait.chr and this_trait.mb: this_trait.location_repr = 'Chr%s: %.6f' % ( this_trait.chr, float(this_trait.mb)) # Get mean expression value query = ( """select ProbeSetXRef.mean from ProbeSetXRef, ProbeSet where ProbeSetXRef.ProbeSetFreezeId = %s and ProbeSet.Id = ProbeSetXRef.ProbeSetId and ProbeSet.Name = '%s' """ % (escape(str(this_trait.dataset.id)), escape(this_trait.name))) #logger.debug("query is:", pf(query)) logger.sql(query) result = g.db.execute(query).fetchone() mean = result[0] if result else 0 if mean: this_trait.mean = "%2.3f" % mean # LRS and its location this_trait.LRS_score_repr = 'N/A' this_trait.LRS_location_repr = 'N/A' # Max LRS and its Locus location if this_trait.lrs and this_trait.locus: query = """ select Geno.Chr, Geno.Mb from Geno, Species where Species.Name = '{}' and Geno.Name = '{}' and Geno.SpeciesId = Species.Id """.format(species, this_trait.locus) logger.sql(query) result = g.db.execute(query).fetchone() if result: lrs_chr, lrs_mb = result this_trait.LRS_score_repr = '%3.1f' % this_trait.lrs this_trait.LRS_location_repr = 'Chr%s: %.6f' % ( lrs_chr, float(lrs_mb)) return trait_list def retrieve_sample_data(self, trait): query = """ SELECT Strain.Name, ProbeSetData.value, ProbeSetSE.error, NStrain.count, Strain.Name2 FROM (ProbeSetData, ProbeSetFreeze, Strain, ProbeSet, ProbeSetXRef) left join ProbeSetSE on (ProbeSetSE.DataId = ProbeSetData.Id AND ProbeSetSE.StrainId = ProbeSetData.StrainId) left join NStrain on (NStrain.DataId = ProbeSetData.Id AND NStrain.StrainId = ProbeSetData.StrainId) WHERE ProbeSet.Name = '%s' AND ProbeSetXRef.ProbeSetId = ProbeSet.Id AND ProbeSetXRef.ProbeSetFreezeId = ProbeSetFreeze.Id AND ProbeSetFreeze.Name = '%s' AND ProbeSetXRef.DataId = ProbeSetData.Id AND ProbeSetData.StrainId = Strain.Id Order BY Strain.Name """ % (escape(trait), escape(self.name)) # logger.sql(query) results = g.db.execute(query).fetchall() #logger.debug("RETRIEVED RESULTS HERE:", results) return results def retrieve_genes(self, column_name): query = """ select ProbeSet.Name, ProbeSet.%s from ProbeSet,ProbeSetXRef where ProbeSetXRef.ProbeSetFreezeId = %s and ProbeSetXRef.ProbeSetId=ProbeSet.Id; """ % (column_name, escape(str(self.id))) # logger.sql(query) results = g.db.execute(query).fetchall() return dict(results) class TempDataSet(DataSet): '''Temporary user-generated data set''' DS_NAME_MAP['Temp'] = 'TempDataSet' def setup(self): self.search_fields = ['name', 'description'] self.display_fields = ['name', 'description'] self.header_fields = ['Name', 'Description'] self.type = 'Temp' # Need to double check later how these are used self.id = 1 self.fullname = 'Temporary Storage' self.shortname = 'Temp' class PhenotypeDataSet(DataSet): DS_NAME_MAP['Publish'] = 'PhenotypeDataSet' def setup(self): #logger.debug("IS A PHENOTYPEDATASET") # Fields in the database table self.search_fields = ['Phenotype.Post_publication_description', 'Phenotype.Pre_publication_description', 'Phenotype.Pre_publication_abbreviation', 'Phenotype.Post_publication_abbreviation', 'PublishXRef.mean', 'Phenotype.Lab_code', 'Publication.PubMed_ID', 'Publication.Abstract', 'Publication.Title', 'Publication.Authors', 'PublishXRef.Id'] # Figure out what display_fields is self.display_fields = ['name', 'group_code', 'pubmed_id', 'pre_publication_description', 'post_publication_description', 'original_description', 'pre_publication_abbreviation', 'post_publication_abbreviation', 'mean', 'lab_code', 'submitter', 'owner', 'authorized_users', 'authors', 'title', 'abstract', 'journal', 'volume', 'pages', 'month', 'year', 'sequence', 'units', 'comments'] # Fields displayed in the search results table header self.header_fields = ['Index', 'Record', 'Description', 'Authors', 'Year', 'Max LRS', 'Max LRS Location', 'Additive Effect'] self.type = 'Publish' self.query_for_group = ''' SELECT InbredSet.Name, InbredSet.Id, InbredSet.GeneticType FROM InbredSet, PublishFreeze WHERE PublishFreeze.InbredSetId = InbredSet.Id AND PublishFreeze.Name = "%s" ''' % escape(self.name) def check_confidentiality(self): # (Urgently?) Need to write this pass def get_trait_info(self, trait_list, species=''): for this_trait in trait_list: if not this_trait.haveinfo: this_trait.retrieve_info(get_qtl_info=True) description = this_trait.post_publication_description # If the dataset is confidential and the user has access to confidential # phenotype traits, then display the pre-publication description instead # of the post-publication description if this_trait.confidential: this_trait.description_display = "" continue # todo for now, because no authorization features if not webqtlUtil.has_access_to_confidentail_phenotype_trait( privilege=self.privilege, userName=self.userName, authorized_users=this_trait.authorized_users): description = this_trait.pre_publication_description if len(description) > 0: this_trait.description_display = description.strip() else: this_trait.description_display = "" if not this_trait.year.isdigit(): this_trait.pubmed_text = "N/A" else: this_trait.pubmed_text = this_trait.year if this_trait.pubmed_id: this_trait.pubmed_link = webqtlConfig.PUBMEDLINK_URL % this_trait.pubmed_id # LRS and its location this_trait.LRS_score_repr = "N/A" this_trait.LRS_location_repr = "N/A" if this_trait.lrs: query = """ select Geno.Chr, Geno.Mb from Geno, Species where Species.Name = '%s' and Geno.Name = '%s' and Geno.SpeciesId = Species.Id """ % (species, this_trait.locus) result = g.db.execute(query).fetchone() if result: if result[0] and result[1]: LRS_Chr = result[0] LRS_Mb = result[1] this_trait.LRS_score_repr = LRS_score_repr = '%3.1f' % this_trait.lrs this_trait.LRS_location_repr = LRS_location_repr = 'Chr%s: %.6f' % ( LRS_Chr, float(LRS_Mb)) def retrieve_sample_data(self, trait): query = """ SELECT Strain.Name, PublishData.value, PublishSE.error, NStrain.count, Strain.Name2 FROM (PublishData, Strain, PublishXRef, PublishFreeze) left join PublishSE on (PublishSE.DataId = PublishData.Id AND PublishSE.StrainId = PublishData.StrainId) left join NStrain on (NStrain.DataId = PublishData.Id AND NStrain.StrainId = PublishData.StrainId) WHERE PublishXRef.InbredSetId = PublishFreeze.InbredSetId AND PublishData.Id = PublishXRef.DataId AND PublishXRef.Id = %s AND PublishFreeze.Id = %s AND PublishData.StrainId = Strain.Id Order BY Strain.Name """ results = g.db.execute(query, (trait, self.id)).fetchall() return results class GenotypeDataSet(DataSet): DS_NAME_MAP['Geno'] = 'GenotypeDataSet' def setup(self): # Fields in the database table self.search_fields = ['Name', 'Chr'] # Find out what display_fields is self.display_fields = ['name', 'chr', 'mb', 'source2', 'sequence'] # Fields displayed in the search results table header self.header_fields = ['Index', 'ID', 'Location'] # Todo: Obsolete or rename this field self.type = 'Geno' self.query_for_group = ''' SELECT InbredSet.Name, InbredSet.Id, InbredSet.GeneticType FROM InbredSet, GenoFreeze WHERE GenoFreeze.InbredSetId = InbredSet.Id AND GenoFreeze.Name = "%s" ''' % escape(self.name) def check_confidentiality(self): return geno_mrna_confidentiality(self) def get_trait_info(self, trait_list, species=None): for this_trait in trait_list: if not this_trait.haveinfo: this_trait.retrieveInfo() if this_trait.chr and this_trait.mb: this_trait.location_repr = 'Chr%s: %.6f' % ( this_trait.chr, float(this_trait.mb)) def retrieve_sample_data(self, trait): query = """ SELECT Strain.Name, GenoData.value, GenoSE.error, "N/A", Strain.Name2 FROM (GenoData, GenoFreeze, Strain, Geno, GenoXRef) left join GenoSE on (GenoSE.DataId = GenoData.Id AND GenoSE.StrainId = GenoData.StrainId) WHERE Geno.SpeciesId = %s AND Geno.Name = %s AND GenoXRef.GenoId = Geno.Id AND GenoXRef.GenoFreezeId = GenoFreeze.Id AND GenoFreeze.Name = %s AND GenoXRef.DataId = GenoData.Id AND GenoData.StrainId = Strain.Id Order BY Strain.Name """ results = g.db.execute(query, (webqtlDatabaseFunction.retrieve_species_id(self.group.name), trait, self.name)).fetchall() return results def geno_mrna_confidentiality(ob): dataset_table = ob.type + "Freeze" #logger.debug("dataset_table [%s]: %s" % (type(dataset_table), dataset_table)) query = '''SELECT Id, Name, FullName, confidentiality, AuthorisedUsers FROM %s WHERE Name = "%s"''' % (dataset_table, ob.name) # result = g.db.execute(query) (_dataset_id, _name, _full_name, confidential, _authorized_users) = result.fetchall()[0] if confidential: return True