diff options
Diffstat (limited to 'wqflask/base/data_set.py')
| -rw-r--r-- | wqflask/base/data_set.py | 683 |
1 files changed, 350 insertions, 333 deletions
diff --git a/wqflask/base/data_set.py b/wqflask/base/data_set.py index 1457ba8d..8906ab69 100644 --- a/wqflask/base/data_set.py +++ b/wqflask/base/data_set.py @@ -17,8 +17,28 @@ # at rwilliams@uthsc.edu and xzhou15@uthsc.edu # # This module is used by GeneNetwork project (www.genenetwork.org) - -from __future__ import absolute_import, print_function, division +from dataclasses import dataclass +from dataclasses import field +from dataclasses import InitVar +from typing import Optional, Dict +from db.call import fetchall, fetchone, fetch1 +from utility.logger import getLogger +from utility.tools import USE_GN_SERVER, USE_REDIS, flat_files, flat_file_exists, GN2_BASE_URL +from db.gn_server import menu_main +from pprint import pformat as pf +from utility.db_tools import escape +from utility.db_tools import mescape +from utility.db_tools import create_in_clause +from maintenance import get_group_samplelists +from utility.tools import locate, locate_ignore_error, flat_files +from utility import gen_geno_ob +from utility import chunks +from utility.benchmark import Bench +from utility import webqtlUtil +from db import webqtlDatabaseFunction +from base import species +from base import webqtlConfig +from flask import Flask, g import os import math import string @@ -28,43 +48,25 @@ import codecs import json import requests import gzip -import cPickle as pickle +import pickle as pickle import itertools from redis import Redis -Redis = Redis() -from flask import Flask, g +r = Redis() -import reaper - -from base import webqtlConfig -from base import species -from db import webqtlDatabaseFunction -from utility import webqtlUtil -from utility.benchmark import Bench -from utility import chunks -from utility import gen_geno_ob -from utility.tools import locate, locate_ignore_error, flat_files - -from wqflask.api import gen_menu - -from maintenance import get_group_samplelists - -from MySQLdb import escape_string as escape -from pprint import pformat as pf -from db.gn_server import menu_main -from db.call import fetchall,fetchone,fetch1 - -from utility.tools import USE_GN_SERVER, USE_REDIS, flat_files, flat_file_exists, GN2_BASE_URL -from utility.logger import getLogger -logger = getLogger(__name__ ) +logger = getLogger(__name__) # Used by create_database to instantiate objects # Each subclass will add to this DS_NAME_MAP = {} -def create_dataset(dataset_name, dataset_type = None, get_samplelist = True, group_name = None): + +def create_dataset(dataset_name, dataset_type=None, + get_samplelist=True, group_name=None): + if dataset_name == "Temp": + dataset_type = "Temp" + if not dataset_type: dataset_type = Dataset_Getter(dataset_name) @@ -75,11 +77,11 @@ def create_dataset(dataset_name, dataset_type = None, get_samplelist = True, gro else: return dataset_class(dataset_name, get_samplelist) -class Dataset_Types(object): - def __init__(self): - """Create a dictionary of samples where the value is set to Geno, -Publish or ProbeSet. E.g. +@dataclass +class DatasetType: + """Create a dictionary of samples where the value is set to Geno, + Publish or ProbeSet. E.g. {'AD-cases-controls-MyersGeno': 'Geno', 'AD-cases-controls-MyersPublish': 'Publish', @@ -90,20 +92,28 @@ Publish or ProbeSet. E.g. 'All Phenotypes': 'Publish', 'B139_K_1206_M': 'ProbeSet', 'B139_K_1206_R': 'ProbeSet' ... - + } """ - self.datasets = {} + redis_instance: InitVar[Redis] + datasets: Optional[Dict] = field(init=False, default_factory=dict) + data: Optional[Dict] = field(init=False) - data = Redis.get("dataset_structure") + def __post_init__(self, redis_instance): + self.redis_instance = redis_instance + data = redis_instance.get("dataset_structure") if data: self.datasets = json.loads(data) - else: #ZS: I don't think this should ever run unless Redis is emptied + else: + # ZS: I don't think this should ever run unless Redis is + # emptied try: - data = json.loads(requests.get(GN2_BASE_URL + "/api/v_pre1/gen_dropdown", timeout = 5).content) - 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]: + data = json.loads(requests.get( + GN2_BASE_URL + "/api/v_pre1/gen_dropdown", + timeout=5).content) + 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" @@ -112,84 +122,77 @@ Publish or ProbeSet. E.g. else: new_type = "ProbeSet" self.datasets[short_dataset_name] = new_type - except: + except Exception: # Do nothing pass - Redis.set("dataset_structure", json.dumps(self.datasets)) + self.redis_instance.set("dataset_structure", + json.dumps(self.datasets)) + self.data = data - # Set LOG_LEVEL_DEBUG=5 to see the following: - logger.debugf(5, "datasets",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 - def __call__(self, name): - if name not in self.datasets: - mrna_expr_query = """ - SELECT - ProbeSetFreeze.Id - FROM - ProbeSetFreeze - WHERE - ProbeSetFreeze.Name = "{0}" - """.format(name) + """ + sql_query_mapping = { + 'mrna_expr': ("SELECT ProbeSetFreeze.Id FROM " + "ProbeSetFreeze WHERE " + "ProbeSetFreeze.Name = \"%s\" "), + 'pheno': ("SELECT InfoFiles.GN_AccesionId " + "FROM InfoFiles, PublishFreeze, InbredSet " + "WHERE InbredSet.Name = '%s' AND " + "PublishFreeze.InbredSetId = InbredSet.Id AND " + "InfoFiles.InfoPageName = PublishFreeze.Name"), + 'other_pheno': ("SELECT PublishFreeze.Name " + "FROM PublishFreeze, InbredSet " + "WHERE InbredSet.Name = '%s' AND " + "PublishFreeze.InbredSetId = InbredSet.Id"), + 'geno': ("SELECT GenoFreeze.Id FROM GenoFreeze WHERE " + "GenoFreeze.Name = \"%s\" ") + } + + 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(mrna_expr_query).fetchall() - if len(results): - self.datasets[name] = "ProbeSet" - Redis.set("dataset_structure", json.dumps(self.datasets)) - return self.datasets[name] + results = g.db.execute(sql_query_mapping[t] % 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 - group_name = name.replace("Publish", "") + def __call__(self, name): + if name not in self.datasets: + for t in ["mrna_expr", "pheno", "other_pheno", "geno"]: + # This has side-effects, with the end result being a + # truth-y value + if(self.set_dataset_key(t, name)): + break + # Return None if name has not been set + return self.datasets.get(name, None) - pheno_query = """SELECT InfoFiles.GN_AccesionId - FROM InfoFiles, PublishFreeze, InbredSet - WHERE InbredSet.Name = '{0}' AND - PublishFreeze.InbredSetId = InbredSet.Id AND - InfoFiles.InfoPageName = PublishFreeze.Name""".format(group_name) - - results = g.db.execute(pheno_query).fetchall() - if len(results): - self.datasets[name] = "Publish" - Redis.set("dataset_structure", json.dumps(self.datasets)) - return self.datasets[name] - - #ZS: For when there isn't an InfoFiles ID; not sure if this and the preceding query are both necessary - other_pheno_query = """SELECT PublishFreeze.Name - FROM PublishFreeze, InbredSet - WHERE InbredSet.Name = '{}' AND - PublishFreeze.InbredSetId = InbredSet.Id""".format(group_name) - - results = g.db.execute(other_pheno_query).fetchall() - if len(results): - self.datasets[name] = "Publish" - Redis.set("dataset_structure", json.dumps(self.datasets)) - return self.datasets[name] - - geno_query = """ - SELECT - GenoFreeze.Id - FROM - GenoFreeze - WHERE - GenoFreeze.Name = "{0}" - """.format(name) - - results = g.db.execute(geno_query).fetchall() - if len(results): - self.datasets[name] = "Geno" - Redis.set("dataset_structure", json.dumps(self.datasets)) - return self.datasets[name] - - #ZS: It shouldn't ever reach this - return None - else: - return self.datasets[name] # Do the intensive work at startup one time only -Dataset_Getter = Dataset_Types() +Dataset_Getter = DatasetType(r) + def create_datasets_list(): if USE_REDIS: key = "all_datasets" - result = Redis.get(key) + result = r.get(key) if result: logger.debug("Redis cache hit") @@ -205,38 +208,25 @@ def create_datasets_list(): for dataset_type in type_dict: query = "SELECT Name FROM {}".format(type_dict[dataset_type]) for result in fetchall(query): - #The query at the beginning of this function isn't - #necessary here, but still would rather just reuse - #it logger.debug("type: {}\tname: - #{}".format(dataset_type, result.Name)) + # The query at the beginning of this function isn't + # necessary here, but still would rather just reuse + # it logger.debug("type: {}\tname: + # {}".format(dataset_type, result.Name)) dataset = create_dataset(result.Name, dataset_type) datasets.append(dataset) if USE_REDIS: - Redis.set(key, pickle.dumps(datasets, pickle.HIGHEST_PROTOCOL)) - Redis.expire(key, 60*60) + r.set(key, pickle.dumps(datasets, pickle.HIGHEST_PROTOCOL)) + r.expire(key, 60 * 60) return datasets -def create_in_clause(items): - """Create an in clause for mysql""" - in_clause = ', '.join("'{}'".format(x) for x in mescape(*items)) - in_clause = '( {} )'.format(in_clause) - return in_clause - - -def mescape(*items): - """Multiple escape""" - escaped = [escape(str(item)) for item in items] - #logger.debug("escaped is:", escaped) - return escaped - - -class Markers(object): +class Markers: """Todo: Build in cacheing so it saves us reading the same file more than once""" + def __init__(self, name): - json_data_fh = open(locate(name + ".json",'genotype/json')) + json_data_fh = open(locate(name + ".json", 'genotype/json')) markers = [] with open("%s/%s_snps.txt" % (flat_files('genotype/bimbam'), name), 'r') as bimbam_fh: @@ -251,7 +241,8 @@ class Markers(object): for line in bimbam_fh: marker = {} marker['name'] = line.split(delimiter)[0].rstrip() - marker['Mb'] = float(line.split(delimiter)[1].rstrip())/1000000 + marker['Mb'] = float(line.split(delimiter)[ + 1].rstrip()) / 1000000 marker['chr'] = line.split(delimiter)[2].rstrip() markers.append(marker) @@ -266,12 +257,12 @@ class Markers(object): logger.debug("length of self.markers:", len(self.markers)) logger.debug("length of p_values:", len(p_values)) - if type(p_values) is list: + if isinstance(p_values, list): # THIS IS only needed for the case when we are limiting the number of p-values calculated - #if len(self.markers) > len(p_values): + # if len(self.markers) > len(p_values): # self.markers = self.markers[:len(p_values)] - for marker, p_value in itertools.izip(self.markers, p_values): + for marker, p_value in zip(self.markers, p_values): if not p_value: continue marker['p_value'] = float(p_value) @@ -280,38 +271,32 @@ class Markers(object): marker['lrs_value'] = 0 else: marker['lod_score'] = -math.log10(marker['p_value']) - #Using -log(p) for the LRS; need to ask Rob how he wants to get LRS from p-values + # Using -log(p) for the LRS; need to ask Rob how he wants to get LRS from p-values marker['lrs_value'] = -math.log10(marker['p_value']) * 4.61 - elif type(p_values) is dict: + elif isinstance(p_values, dict): filtered_markers = [] for marker in self.markers: - #logger.debug("marker[name]", marker['name']) - #logger.debug("p_values:", p_values) if marker['name'] in p_values: - #logger.debug("marker {} IS in p_values".format(i)) marker['p_value'] = p_values[marker['name']] if math.isnan(marker['p_value']) or (marker['p_value'] <= 0): marker['lod_score'] = 0 marker['lrs_value'] = 0 else: marker['lod_score'] = -math.log10(marker['p_value']) - #Using -log(p) for the LRS; need to ask Rob how he wants to get LRS from p-values - marker['lrs_value'] = -math.log10(marker['p_value']) * 4.61 + # Using -log(p) for the LRS; need to ask Rob how he wants to get LRS from p-values + marker['lrs_value'] = - \ + math.log10(marker['p_value']) * 4.61 filtered_markers.append(marker) - #else: - #logger.debug("marker {} NOT in p_values".format(i)) - #self.markers.remove(marker) - #del self.markers[i] self.markers = filtered_markers + class HumanMarkers(Markers): - def __init__(self, name, specified_markers = []): + def __init__(self, name, specified_markers=[]): marker_data_fh = open(flat_files('mapping') + '/' + name + '.bim') self.markers = [] for line in marker_data_fh: splat = line.strip().split() - #logger.debug("splat:", splat) if len(specified_markers) > 0: if splat[1] in specified_markers: marker = {} @@ -327,14 +312,11 @@ class HumanMarkers(Markers): marker['Mb'] = float(splat[3]) / 1000000 self.markers.append(marker) - #logger.debug("markers is: ", pf(self.markers)) - - def add_pvalues(self, p_values): super(HumanMarkers, self).add_pvalues(p_values) -class DatasetGroup(object): +class DatasetGroup: """ Each group has multiple datasets; each species has multiple groups. @@ -342,12 +324,15 @@ class DatasetGroup(object): 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) + 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) + 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" @@ -366,7 +351,8 @@ class DatasetGroup(object): def get_mapping_methods(self): - mapping_id = g.db.execute("select MappingMethodId from InbredSet where Name= '%s'" % self.name).fetchone()[0] + 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": @@ -383,8 +369,8 @@ class DatasetGroup(object): def get_markers(self): def check_plink_gemma(): if flat_file_exists("mapping"): - MAPPING_PATH = flat_files("mapping")+"/" - if os.path.isfile(MAPPING_PATH+self.name+".bed"): + MAPPING_PATH = flat_files("mapping") + "/" + if os.path.isfile(MAPPING_PATH + self.name + ".bed"): return True return False @@ -410,6 +396,15 @@ class DatasetGroup(object): if maternal and paternal: self.parlist = [maternal, paternal] + def get_study_samplelists(self): + study_sample_file = locate_ignore_error(self.name + ".json", 'study_sample_lists') + try: + f = open(study_sample_file) + except: + return [] + study_samples = json.load(f) + return study_samples + def get_genofiles(self): jsonfile = "%s/%s.json" % (webqtlConfig.GENODIR, self.name) try: @@ -423,22 +418,23 @@ class DatasetGroup(object): result = None key = "samplelist:v3:" + self.name if USE_REDIS: - result = Redis.get(key) + result = r.get(key) if result is not None: self.samplelist = json.loads(result) else: logger.debug("Cache not hit") - genotype_fn = locate_ignore_error(self.name+".geno",'genotype') + genotype_fn = locate_ignore_error(self.name + ".geno", 'genotype') if genotype_fn: - self.samplelist = get_group_samplelists.get_samplelist("geno", genotype_fn) + self.samplelist = get_group_samplelists.get_samplelist( + "geno", genotype_fn) else: self.samplelist = None if USE_REDIS: - Redis.set(key, json.dumps(self.samplelist)) - Redis.expire(key, 60*5) + r.set(key, json.dumps(self.samplelist)) + r.expire(key, 60 * 5) def all_samples_ordered(self): result = [] @@ -448,32 +444,28 @@ class DatasetGroup(object): def read_genotype_file(self, use_reaper=False): '''Read genotype from .geno file instead of database''' - #genotype_1 is Dataset Object without parents and f1 - #genotype_2 is Dataset Object with parents and f1 (not for intercross) + # genotype_1 is Dataset Object without parents and f1 + # genotype_2 is Dataset Object with parents and f1 (not for intercross) - #genotype_1 = reaper.Dataset() # reaper barfs on unicode filenames, so here we ensure it's a string if self.genofile: - if "RData" in self.genofile: #ZS: This is a temporary fix; I need to change the way the JSON files that point to multiple genotype files are structured to point to other file types like RData - full_filename = str(locate(self.genofile.split(".")[0] + ".geno", 'genotype')) + if "RData" in self.genofile: # ZS: This is a temporary fix; I need to change the way the JSON files that point to multiple genotype files are structured to point to other file types like RData + full_filename = str( + locate(self.genofile.split(".")[0] + ".geno", 'genotype')) else: full_filename = str(locate(self.genofile, 'genotype')) else: full_filename = str(locate(self.name + '.geno', 'genotype')) - - if use_reaper: - genotype_1 = reaper.Dataset() - genotype_1.read(full_filename) - else: - genotype_1 = gen_geno_ob.genotype(full_filename) + genotype_1 = gen_geno_ob.genotype(full_filename) if genotype_1.type == "group" and self.parlist: - genotype_2 = genotype_1.add(Mat=self.parlist[0], Pat=self.parlist[1]) #, F1=_f1) + genotype_2 = genotype_1.add( + Mat=self.parlist[0], Pat=self.parlist[1]) # , F1=_f1) else: genotype_2 = genotype_1 - #determine default genotype object + # determine default genotype object if self.incparentsf1 and genotype_1.type != "intercross": genotype = genotype_2 else: @@ -484,27 +476,21 @@ class DatasetGroup(object): return genotype -def datasets(group_name, this_group = None): + +def datasets(group_name, this_group=None): key = "group_dataset_menu:v2:" + group_name - logger.debug("key is2:", key) dataset_menu = [] - logger.debug("[tape4] webqtlConfig.PUBLICTHRESH:", webqtlConfig.PUBLICTHRESH) - logger.debug("[tape4] type webqtlConfig.PUBLICTHRESH:", type(webqtlConfig.PUBLICTHRESH)) the_results = fetchall(''' (SELECT '#PublishFreeze',PublishFreeze.FullName,PublishFreeze.Name FROM PublishFreeze,InbredSet WHERE PublishFreeze.InbredSetId = InbredSet.Id and InbredSet.Name = '%s' - and PublishFreeze.public > %s - and PublishFreeze.confidentiality < 1 ORDER BY PublishFreeze.Id ASC) UNION (SELECT '#GenoFreeze',GenoFreeze.FullName,GenoFreeze.Name FROM GenoFreeze, InbredSet WHERE GenoFreeze.InbredSetId = InbredSet.Id - and InbredSet.Name = '%s' - and GenoFreeze.public > %s - and GenoFreeze.confidentiality < 1) + and InbredSet.Name = '%s') UNION (SELECT Tissue.Name, ProbeSetFreeze.FullName,ProbeSetFreeze.Name FROM ProbeSetFreeze, ProbeFreeze, InbredSet, Tissue @@ -512,16 +498,15 @@ def datasets(group_name, this_group = None): and ProbeFreeze.TissueId = Tissue.Id and ProbeFreeze.InbredSetId = InbredSet.Id and InbredSet.Name like %s - and ProbeSetFreeze.public > %s - and ProbeSetFreeze.confidentiality < 1 ORDER BY Tissue.Name, ProbeSetFreeze.OrderList DESC) - ''' % (group_name, webqtlConfig.PUBLICTHRESH, - group_name, webqtlConfig.PUBLICTHRESH, - "'" + group_name + "'", webqtlConfig.PUBLICTHRESH)) + ''' % (group_name, + group_name, + "'" + group_name + "'")) sorted_results = sorted(the_results, key=lambda kv: kv[0]) - pheno_inserted = False #ZS: This is kind of awkward, but need to ensure Phenotypes show up before Genotypes in dropdown + # ZS: This is kind of awkward, but need to ensure Phenotypes show up before Genotypes in dropdown + pheno_inserted = False geno_inserted = False for dataset_item in sorted_results: tissue_name = dataset_item[0] @@ -529,13 +514,16 @@ def datasets(group_name, this_group = None): dataset_short = dataset_item[2] if tissue_name in ['#PublishFreeze', '#GenoFreeze']: if tissue_name == '#PublishFreeze' and (dataset_short == group_name + 'Publish'): - dataset_menu.insert(0, dict(tissue=None, datasets=[(dataset, dataset_short)])) + dataset_menu.insert( + 0, dict(tissue=None, datasets=[(dataset, dataset_short)])) pheno_inserted = True elif pheno_inserted and tissue_name == '#GenoFreeze': - dataset_menu.insert(1, dict(tissue=None, datasets=[(dataset, dataset_short)])) + dataset_menu.insert( + 1, dict(tissue=None, datasets=[(dataset, dataset_short)])) geno_inserted = True else: - dataset_menu.append(dict(tissue=None, datasets=[(dataset, dataset_short)])) + dataset_menu.append( + dict(tissue=None, datasets=[(dataset, dataset_short)])) else: tissue_already_exists = False for i, tissue_dict in enumerate(dataset_menu): @@ -544,15 +532,14 @@ def datasets(group_name, this_group = None): break if tissue_already_exists: - #logger.debug("dataset_menu:", dataset_menu[i]['datasets']) dataset_menu[i]['datasets'].append((dataset, dataset_short)) else: dataset_menu.append(dict(tissue=tissue_name, - datasets=[(dataset, dataset_short)])) + datasets=[(dataset, dataset_short)])) if USE_REDIS: - Redis.set(key, pickle.dumps(dataset_menu, pickle.HIGHEST_PROTOCOL)) - Redis.expire(key, 60*5) + r.set(key, pickle.dumps(dataset_menu, pickle.HIGHEST_PROTOCOL)) + r.expire(key, 60 * 5) if this_group != None: this_group._datasets = dataset_menu @@ -560,14 +547,15 @@ def datasets(group_name, this_group = None): else: return dataset_menu -class DataSet(object): + +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): + def __init__(self, name, get_samplelist=True, group_name=None): assert name, "Need a name" self.name = name @@ -575,30 +563,34 @@ class DataSet(object): self.shortname = None self.fullname = None self.type = None - self.data_scale = None #ZS: For example log2 + self.data_scale = None # ZS: For example log2 + self.accession_id = None self.setup() - if self.type == "Temp": #Need to supply group name as input if temp trait - self.group = DatasetGroup(self, name=group_name) # sets self.group and self.group_id and gets genotype + 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() - self.group = DatasetGroup(self) # sets self.group and self.group_id and gets genotype + # 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.group.get_samplelist() self.species = species.TheSpecies(self) - - def get_desc(self): - """Gets overridden later, at least for Temp...used by trait's get_given_name""" - return None - - # Delete this eventually - @property - def riset(): - Weve_Renamed_This_As_Group + def as_dict(self): + return { + 'name': self.name, + 'shortname': self.shortname, + 'fullname': self.fullname, + 'type': self.type, + 'data_scale': self.data_scale, + 'group': self.group.name, + 'accession_id': self.accession_id + } def get_accession_id(self): if self.type == "Publish": @@ -637,29 +629,26 @@ class DataSet(object): """ - try: if self.type == "ProbeSet": query_args = tuple(escape(x) for x in ( - str(webqtlConfig.PUBLICTHRESH), 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.public > %s -AND 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"]) + 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"), - str(webqtlConfig.PUBLICTHRESH), self.name, self.name, self.name)) @@ -668,14 +657,77 @@ AND (ProbeSetFreeze.Name = '%s' OR ProbeSetFreeze.FullName = '%s' OR ProbeSetFre self.id, self.name, self.fullname, self.shortname = fetchone(""" SELECT Id, Name, FullName, ShortName FROM %s - WHERE public > %s AND - (Name = '%s' OR FullName = '%s' OR ShortName = '%s') - """ % (query_args)) + WHERE (Name = '%s' OR FullName = '%s' OR ShortName = '%s') + """ % (query_args)) except TypeError: - logger.debug("Dataset {} is not yet available in GeneNetwork.".format(self.name)) + logger.debug( + "Dataset {} is not yet available in GeneNetwork.".format(self.name)) pass + def chunk_dataset(self, dataset, n): + + results = {} + + query = """ + SELECT ProbeSetXRef.DataId,ProbeSet.Name + FROM ProbeSet, ProbeSetXRef, ProbeSetFreeze + WHERE ProbeSetFreeze.Name = '{}' AND + ProbeSetXRef.ProbeSetFreezeId = ProbeSetFreeze.Id AND + ProbeSetXRef.ProbeSetId = ProbeSet.Id + """.format(self.name) + + # should cache this + + traits_name_dict = dict(g.db.execute(query).fetchall()) + + for i in range(0, len(dataset), n): + matrix = list(dataset[i:i + n]) + trait_name = traits_name_dict[matrix[0][0]] + + my_values = [value for (trait_name, strain, value) in matrix] + results[trait_name] = my_values + return results + + def get_probeset_data(self, sample_list=None, trait_ids=None): + + # improvement of get trait data--->>> + 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)) + results = dict(g.db.execute(query).fetchall()) + sample_ids = [results[item] for item in self.samplelist] + + sorted_samplelist = [strain_name for strain_name, strain_id in sorted( + results.items(), key=lambda item: item[1])] + + query = """SELECT * from ProbeSetData + where StrainID in {} + and id in (SELECT ProbeSetXRef.DataId + FROM (ProbeSet, ProbeSetXRef, ProbeSetFreeze) + WHERE ProbeSetXRef.ProbeSetFreezeId = ProbeSetFreeze.Id + and ProbeSetFreeze.Name = '{}' + and ProbeSet.Id = ProbeSetXRef.ProbeSetId)""".format(create_in_clause(sample_ids), self.name) + + query_results = list(g.db.execute(query).fetchall()) + data_results = self.chunk_dataset(query_results, len(sample_ids)) + self.samplelist = sorted_samplelist + self.trait_data = data_results + + def get_trait_data(self, sample_list=None): if sample_list: self.samplelist = sample_list @@ -692,7 +744,6 @@ AND (ProbeSetFreeze.Name = '%s' OR ProbeSetFreeze.FullName = '%s' OR ProbeSetFre 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] @@ -713,7 +764,7 @@ AND (ProbeSetFreeze.Name = '%s' OR ProbeSetFreeze.FullName = '%s' OR ProbeSetFre else: query = "SELECT {}.Name,".format(escape(dataset_type)) data_start_pos = 1 - query += string.join(temp, ', ') + query += ', '.join(temp) query += ' FROM ({}, {}XRef, {}Freeze) '.format(*mescape(dataset_type, self.type, self.type)) @@ -731,7 +782,7 @@ AND (ProbeSetFreeze.Name = '%s' OR ProbeSetFreeze.FullName = '%s' OR ProbeSetFre 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)) + dataset_type, self.type, dataset_type, dataset_type)) else: query += """ WHERE {}XRef.{}FreezeId = {}Freeze.Id @@ -739,13 +790,9 @@ AND (ProbeSetFreeze.Name = '%s' OR ProbeSetFreeze.FullName = '%s' OR ProbeSetFre 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)) + self.name, dataset_type, self.type, self.type, dataset_type)) - #logger.debug("trait data query: ", query) - - logger.sql(query) results = g.db.execute(query).fetchall() - #logger.debug("query results:", results) trait_sample_data.append(results) trait_count = len(trait_sample_data[0]) @@ -759,25 +806,23 @@ AND (ProbeSetFreeze.Name = '%s' OR ProbeSetFreeze.FullName = '%s' OR ProbeSetFre self.trait_data[trait_name] += ( trait_sample_data[chunk_counter][trait_counter][data_start_pos:]) + 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'] + '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', @@ -799,13 +844,13 @@ class PhenotypeDataSet(DataSet): # Fields displayed in the search results table header self.header_fields = ['Index', - 'Record', - 'Description', - 'Authors', - 'Year', - 'Max LRS', - 'Max LRS Location', - 'Additive Effect'] + 'Record', + 'Description', + 'Authors', + 'Year', + 'Max LRS', + 'Max LRS Location', + 'Additive Effect'] self.type = 'Publish' @@ -823,7 +868,7 @@ class PhenotypeDataSet(DataSet): # (Urgently?) Need to write this pass - def get_trait_info(self, trait_list, species = ''): + def get_trait_info(self, trait_list, species=''): for this_trait in trait_list: if not this_trait.haveinfo: @@ -831,9 +876,9 @@ class PhenotypeDataSet(DataSet): 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 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 # for now, because no authorization features @@ -858,7 +903,7 @@ class PhenotypeDataSet(DataSet): if this_trait.pubmed_id: this_trait.pubmed_link = webqtlConfig.PUBMEDLINK_URL % this_trait.pubmed_id - #LRS and its location + # LRS and its location this_trait.LRS_score_repr = "N/A" this_trait.LRS_location_repr = "N/A" @@ -869,7 +914,6 @@ class PhenotypeDataSet(DataSet): Geno.Name = '%s' and Geno.SpeciesId = Species.Id """ % (species, this_trait.locus) - logger.sql(query) result = g.db.execute(query).fetchone() if result: @@ -878,7 +922,8 @@ class PhenotypeDataSet(DataSet): 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)) + this_trait.LRS_location_repr = LRS_location_repr = 'Chr%s: %.6f' % ( + LRS_Chr, float(LRS_Mb)) def retrieve_sample_data(self, trait): query = """ @@ -898,7 +943,6 @@ class PhenotypeDataSet(DataSet): Order BY Strain.Name """ - logger.sql(query) results = g.db.execute(query, (trait, self.id)).fetchall() return results @@ -945,12 +989,13 @@ class GenotypeDataSet(DataSet): 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) ) + 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, GenoData.Id, Strain.Name2 + Strain.Name, GenoData.value, GenoSE.error, "N/A", Strain.Name2 FROM (GenoData, GenoFreeze, Strain, Geno, GenoXRef) left join GenoSE on @@ -964,7 +1009,6 @@ class GenotypeDataSet(DataSet): Order BY Strain.Name """ - logger.sql(query) results = g.db.execute(query, (webqtlDatabaseFunction.retrieve_species_id(self.group.name), trait, self.name)).fetchall() @@ -1014,14 +1058,14 @@ class MrnaAssayDataSet(DataSet): # Fields displayed in the search results table header self.header_fields = ['Index', - 'Record', - 'Symbol', - 'Description', - 'Location', - 'Mean', - 'Max LRS', - 'Max LRS Location', - 'Additive Effect'] + 'Record', + 'Symbol', + 'Description', + 'Location', + 'Mean', + 'Max LRS', + 'Max LRS Location', + 'Additive Effect'] # Todo: Obsolete or rename this field self.type = 'ProbeSet' @@ -1037,7 +1081,6 @@ class MrnaAssayDataSet(DataSet): ProbeSetFreeze.Name = "%s" ''' % escape(self.name) - def check_confidentiality(self): return geno_mrna_confidentiality(self) @@ -1055,37 +1098,37 @@ class MrnaAssayDataSet(DataSet): 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 = unicode(str(this_trait.description).strip(codecs.BOM_UTF8), 'utf-8') - target_string = unicode(str(this_trait.probe_target_description).strip(codecs.BOM_UTF8), 'utf-8') + # 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'): + 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)) + this_trait.location_repr = 'Chr%s: %.6f' % ( + this_trait.chr, float(this_trait.mb)) - #Get mean expression value + # Get mean expression value query = ( - """select ProbeSetXRef.mean from ProbeSetXRef, ProbeSet + """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 @@ -1093,11 +1136,11 @@ class MrnaAssayDataSet(DataSet): if mean: this_trait.mean = "%2.3f" % mean - #LRS and its location + # LRS and its location this_trait.LRS_score_repr = 'N/A' this_trait.LRS_location_repr = 'N/A' - #Max LRS and its Locus location + # Max LRS and its Locus location if this_trait.lrs and this_trait.locus: query = """ select Geno.Chr, Geno.Mb from Geno, Species @@ -1105,24 +1148,28 @@ class MrnaAssayDataSet(DataSet): 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)) + 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, ProbeSetData.Id, Strain.Name2 + Strain.Name, ProbeSetData.value, ProbeSetSE.error, NStrain.count, Strain.Name2 FROM - (ProbeSetData, ProbeSetFreeze, Strain, ProbeSet, ProbeSetXRef) + (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 @@ -1132,9 +1179,7 @@ class MrnaAssayDataSet(DataSet): 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): @@ -1144,7 +1189,6 @@ class MrnaAssayDataSet(DataSet): 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) @@ -1173,48 +1217,21 @@ class TempDataSet(DataSet): self.shortname = 'Temp' - @staticmethod - def handle_pca(desc): - if 'PCA' in desc: - # Todo: Modernize below lines - desc = desc[desc.rindex(':')+1:].strip() - else: - desc = desc[:desc.index('entered')].strip() - return desc - - def get_desc(self): - query = 'SELECT description FROM Temp WHERE Name=%s' % self.name - logger.sql(query) - g.db.execute(query) - desc = g.db.fetchone()[0] - desc = self.handle_pca(desc) - return desc - - def retrieve_sample_data(self, trait): - query = """ - SELECT - Strain.Name, TempData.value, TempData.SE, TempData.NStrain, TempData.Id - FROM - TempData, Temp, Strain - WHERE - TempData.StrainId = Strain.Id AND - TempData.Id = Temp.DataId AND - Temp.name = '%s' - Order BY - Strain.Name - """ % escape(trait.name) - - logger.sql(query) - results = g.db.execute(query).fetchall() - - 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) - logger.sql(query) + 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 result = g.db.execute(query) (dataset_id, |