diff options
Diffstat (limited to 'web')
| -rwxr-xr-x | web/webqtl/correlation/CorrelationPage.py | 1143 |
1 files changed, 638 insertions, 505 deletions
diff --git a/web/webqtl/correlation/CorrelationPage.py b/web/webqtl/correlation/CorrelationPage.py index 8ce669cb..42059117 100755 --- a/web/webqtl/correlation/CorrelationPage.py +++ b/web/webqtl/correlation/CorrelationPage.py @@ -23,6 +23,9 @@ # Created by GeneNetwork Core Team 2010/08/10 # # Last updated by NL 2011/02/11 +# Last updated by Christian Fernandez 2012/04/07 +# Refactored correlation calculation into smaller functions in preparation of +# separating html from existing code import string from math import * @@ -47,447 +50,349 @@ from dbFunction import webqtlDatabaseFunction import utility.webqtlUtil #this is for parallel computing only. from correlation import correlationFunction +import logging +logging.basicConfig(filename="/tmp/gn_log", level=logging.INFO) +_log = logging.getLogger("correlation") -class CorrelationPage(templatePage): - - corrMinInformative = 4 - - def __init__(self, fd): - - #XZ, 01/14/2009: This method is for parallel computing only. - #XZ: It is supposed to be called when "Genetic Correlation, Pearson's r" (method 1) - #XZ: or "Genetic Correlation, Spearman's rho" (method 2) is selected - def compute_corr( input_nnCorr, input_trait, input_list, computing_method): - - allcorrelations = [] - - for line in input_list: - tokens = line.split('","') - tokens[-1] = tokens[-1][:-2] #remove the last " - tokens[0] = tokens[0][1:] #remove the first " - - traitdataName = tokens[0] - database_trait = tokens[1:] - - if computing_method == "1": #XZ: Pearson's r - corr,nOverlap = utility.webqtlUtil.calCorrelationText(input_trait, database_trait, input_nnCorr) - else: #XZ: Spearman's rho - corr,nOverlap = utility.webqtlUtil.calCorrelationRankText(input_trait, database_trait, input_nnCorr) - traitinfo = [traitdataName,corr,nOverlap] - allcorrelations.append(traitinfo) - - return allcorrelations - - - templatePage.__init__(self, fd) - - if not self.openMysql(): - return - - if not fd.genotype: - fd.readGenotype() - - #XZ, 09/18/2008: get the information such as value, variance of the input strain names from the form. - if fd.allstrainlist: - mdpchoice = fd.formdata.getvalue('MDPChoice') - #XZ, in HTML source code, it is "BXD Only" or "BXH only", and so on - if mdpchoice == "1": - strainlist = fd.f1list + fd.strainlist - #XZ, in HTML source code, it is "MDP Only" - elif mdpchoice == "2": - strainlist = [] - strainlist2 = fd.f1list + fd.strainlist - for strain in fd.allstrainlist: - if strain not in strainlist2: - strainlist.append(strain) - #So called MDP Panel - if strainlist: - strainlist = fd.f1list+fd.parlist+strainlist - #XZ, in HTML source code, it is "All Cases" - else: - strainlist = fd.allstrainlist - #XZ, 09/18/2008: put the trait data into dictionary fd.allTraitData - fd.readData(fd.allstrainlist) - else: - mdpchoice = None - strainlist = fd.strainlist - #XZ, 09/18/2008: put the trait data into dictionary fd.allTraitData - fd.readData() - - #XZ, 3/16/2010: variable RISet must be pass by the form - RISet = fd.RISet - #XZ, 12/12/2008: get species infomation - species = webqtlDatabaseFunction.retrieveSpecies(cursor=self.cursor, RISet=RISet) - - #XZ, 09/18/2008: get all information about the user selected database. - self.target_db_name = fd.formdata.getvalue('database') - - try: - self.db = webqtlDataset(self.target_db_name, self.cursor) - except: - heading = "Correlation Table" - detail = ["The database you just requested has not been established yet."] - self.error(heading=heading,detail=detail) - return - - #XZ, 09/18/2008: check if user has the authority to get access to the database. - if self.db.type == 'ProbeSet': - self.cursor.execute('SELECT Id, Name, FullName, confidentiality, AuthorisedUsers FROM ProbeSetFreeze WHERE Name = "%s"' % self.target_db_name) - indId, indName, indFullName, confidential, AuthorisedUsers = self.cursor.fetchall()[0] - - if confidential == 1: - access_to_confidential_dataset = 0 - - #for the dataset that confidentiality is 1 - #1. 'admin' and 'root' can see all of the dataset - #2. 'user' can see the dataset that AuthorisedUsers contains his id(stored in the Id field of User table) - if webqtlConfig.USERDICT[self.privilege] > webqtlConfig.USERDICT['user']: - access_to_confidential_dataset = 1 - else: - AuthorisedUsersList=AuthorisedUsers.split(',') - if AuthorisedUsersList.__contains__(self.userName): - access_to_confidential_dataset = 1 - - if not access_to_confidential_dataset: - #Error, Confidential Database - heading = "Correlation Table" - detail = ["The %s database you selected is not open to the public at this time, please go back and select other database." % indFullName] - self.error(heading=heading,detail=detail,error="Confidential Database") - return - - #XZ, 09/18/2008: filter out the strains that have no value. - _strains, _vals, _vars, N = fd.informativeStrains(strainlist) - - N = len(_strains) - - if N < self.corrMinInformative: - heading = "Correlation Table" - detail = ['Fewer than %d strain data were entered for %s data set. No calculation of correlation has been attempted.' % (self.corrMinInformative, RISet)] - self.error(heading=heading,detail=detail) - return - - #XZ, 09/28/2008: if user select "1", then display 1, 3 and 4. - #XZ, 09/28/2008: if user select "2", then display 2, 3 and 5. - #XZ, 09/28/2008: if user select "3", then display 1, 3 and 4. - #XZ, 09/28/2008: if user select "4", then display 1, 3 and 4. - #XZ, 09/28/2008: if user select "5", then display 2, 3 and 5. - methodDict = {"1":"Genetic Correlation (Pearson's r)","2":"Genetic Correlation (Spearman's rho)","3":"SGO Literature Correlation","4":"Tissue Correlation (Pearson's r)", "5":"Tissue Correlation (Spearman's rho)"} - self.method = fd.formdata.getvalue('method') - if self.method not in ("1","2","3","4","5"): - self.method = "1" - - self.returnNumber = int(fd.formdata.getvalue('criteria')) - - myTrait = fd.formdata.getvalue('fullname') - if myTrait: - myTrait = webqtlTrait(fullname=myTrait, cursor=self.cursor) - myTrait.retrieveInfo() - - # We will not get Literature Correlations if there is no GeneId because there is nothing to look against - try: - input_trait_GeneId = int(fd.formdata.getvalue('GeneId')) - except: - input_trait_GeneId = None - - # We will not get Tissue Correlations if there is no gene symbol because there is nothing to look against - try: - input_trait_symbol = myTrait.symbol - except: - input_trait_symbol = None - - - #XZ, 12/12/2008: if the species is rat or human, translate the geneid to mouse geneid - input_trait_mouse_geneid = self.translateToMouseGeneID(species, input_trait_GeneId) - - - #XZ: As of Nov/13/2010, this dataset is 'UTHSC Illumina V6.2 RankInv B6 D2 average CNS GI average (May 08)' - TissueProbeSetFreezeId = 1 - - #XZ, 09/22/2008: If we need search by GeneId, - #XZ, 09/22/2008: we have to check if this GeneId is in the literature or tissue correlation table. - #XZ, 10/15/2008: We also to check if the selected database is probeset type. - if self.method == "3" or self.method == "4" or self.method == "5": - if self.db.type != "ProbeSet": - self.error(heading="Wrong correlation type",detail="It is not possible to compute the %s between your trait and data in this %s database. Please try again after selecting another type of correlation." % (methodDict[self.method],self.db.name),error="Correlation Type Error") - return - - """ - if not input_trait_GeneId: - self.error(heading="No Associated GeneId",detail="This trait has no associated GeneId, so we are not able to show any literature or tissue related information.",error="No GeneId Error") - return - """ - - #XZ: We have checked geneid did exist +debug_file = open("/home/zas1024/gn/web/corr_debug.txt","w") +debug_file2 = open("/home/zas1024/gn/web/corr_debug2.txt","w") - if self.method == "3": - if not input_trait_GeneId or not self.checkForLitInfo(input_trait_mouse_geneid): - self.error(heading="No Literature Info",detail="This gene does not have any associated Literature Information.",error="Literature Correlation Error") - return +METHOD_SAMPLE_PEARSON = "1" +METHOD_SAMPLE_RANK = "2" +METHOD_LIT = "3" +METHOD_TISSUE_PEARSON = "4" +METHOD_TISSUE_RANK = "5" - if self.method == "4" or self.method == "5": - if not input_trait_symbol: - self.error(heading="No Tissue Correlation Information",detail="This gene does not have any associated Tissue Correlation Information.",error="Tissue Correlation Error") - return +TISSUE_METHODS = [METHOD_TISSUE_PEARSON, METHOD_TISSUE_RANK] - if not self.checkSymbolForTissueCorr(TissueProbeSetFreezeId, myTrait.symbol): - self.error(heading="No Tissue Correlation Information",detail="This gene does not have any associated Tissue Correlation Information.",error="Tissue Correlation Error") - return +TISSUE_MOUSE_DB = 1 -############################################################################################################################################ +class AuthException(Exception): pass - allcorrelations = [] - nnCorr = len(_vals) - #XZ: Use the fast method only for probeset dataset, and this dataset must have been created. - #XZ: Otherwise, use original method +class Trait(object): + def __init__(self, name, raw_values = None, lit_corr = None, tissue_corr = None, p_tissue = None): + self.name = name + self.raw_values = raw_values + self.lit_corr = lit_corr + self.tissue_corr = tissue_corr + self.p_tissue = p_tissue + self.correlation = 0 + self.p_value = 0 - useFastMethod = False + @staticmethod + def from_csv(line, data_start = 1): + name = line[0] + numbers = line[data_start:] + # _log.info(numbers) + numbers = [ float(number) for number in numbers ] - if self.db.type == "ProbeSet": + return Trait(name, raw_values = numbers) - DatabaseFileName = self.getFileName( target_db_name=self.target_db_name ) - DirectoryList = os.listdir(webqtlConfig.TEXTDIR) ### List of existing text files. Used to check if a text file already exists - - if DatabaseFileName in DirectoryList: - useFastMethod = True - - if useFastMethod: - if 1: - #try: - useLit = False - if self.method == "3": - litCorrs = self.fetchLitCorrelations(species=species, GeneId=input_trait_GeneId, db=self.db, returnNumber=self.returnNumber) - useLit = True - - useTissueCorr = False - if self.method == "4" or self.method == "5": - tissueCorrs = self.fetchTissueCorrelations(db=self.db, primaryTraitSymbol=input_trait_symbol, TissueProbeSetFreezeId=TissueProbeSetFreezeId, method=self.method, returnNumber = self.returnNumber) - useTissueCorr = True - - datasetFile = open(webqtlConfig.TEXTDIR+DatabaseFileName,'r') - - #XZ, 01/08/2009: read the first line - line = datasetFile.readline() - dataset_strains = webqtlUtil.readLineCSV(line)[1:] - - #XZ, 01/08/2009: This step is critical. It is necessary for this new method. - #XZ: The original function fetchAllDatabaseData uses all strains stored in variable _strains to - #XZ: retrieve the values of each strain from database in real time. - #XZ: The new method uses all strains stored in variable dataset_strains to create a new variable - #XZ: _newvals. _newvals has the same length as dataset_strains. The items in _newvals is in - #XZ: the same order of items in dataset_strains. The value of each item in _newvals is either - #XZ: the value of correspinding strain in _vals or 'None'. - _newvals = [] - for item in dataset_strains: - if item in _strains: - _newvals.append(_vals[_strains.index(item)]) - else: - _newvals.append('None') - - nnCorr = len(_newvals) + def calculate_correlation(self, values, method): + """Calculate the correlation value and p value according to the method specified""" - #XZ, 01/14/2009: If literature corr or tissue corr is selected, - #XZ: there is no need to use parallel computing. - if useLit or useTissueCorr: - for line in datasetFile: - traitdata=webqtlUtil.readLineCSV(line) - traitdataName = traitdata[0] - traitvals = traitdata[1:] + #ZS: This takes the list of values of the trait our selected trait is being correlated against and removes the values of the samples our trait has no value for + #There's probably a better way of dealing with this, but I'll have to ask Christian + updated_raw_values = [] + updated_values = [] + for i in range(len(values)): + if values[i] != "None": + updated_raw_values.append(self.raw_values[i]) + updated_values.append(values[i]) - if useLit: - if not litCorrs.has_key( traitdataName ): - continue + self.raw_values = updated_raw_values + values = updated_values - if useTissueCorr: - if not tissueCorrs.has_key( traitdataName ): - continue - - if self.method == "3" or self.method == "4": - corr,nOverlap = webqtlUtil.calCorrelationText(traitvals,_newvals,nnCorr) - else: - corr,nOverlap = webqtlUtil.calCorrelationRankText(traitvals,_newvals,nnCorr) - - traitinfo = [traitdataName,corr,nOverlap] - - if useLit: - traitinfo.append(litCorrs[traitdataName]) - - if useTissueCorr: - tempCorr, tempPValue = tissueCorrs[traitdataName] - traitinfo.append(tempCorr) - traitinfo.append(tempPValue) + if method == METHOD_SAMPLE_PEARSON or method == METHOD_LIT or method == METHOD_TISSUE_PEARSON: + corr,nOverlap = webqtlUtil.calCorrelation(self.raw_values, values, len(values)) + else: + corr,nOverlap = webqtlUtil.calCorrelationRank(self.raw_values, values, len(values)) - allcorrelations.append(traitinfo) - #XZ, 01/14/2009: If genetic corr is selected, use parallel computing - else: - input_line_list = datasetFile.readlines() - all_line_number = len(input_line_list) + self.correlation = corr + self.overlap = nOverlap - step = 1000 - job_number = math.ceil( float(all_line_number)/step ) + if self.overlap < 3: + self.p_value = 1.0 + else: + #ZS - This is probably the wrong way to deal with this. Correlation values of 1.0 definitely exist (the trait correlated against itself), so zero division needs to br prevented. + if abs(self.correlation) >= 1.0: + self.p_value = 0.0 + else: + ZValue = 0.5*log((1.0+self.correlation)/(1.0-self.correlation)) + ZValue = ZValue*sqrt(self.overlap-3) + self.p_value = 2.0*(1.0 - reaper.normp(abs(ZValue))) - job_input_lists = [] + - for job_index in range( int(job_number) ): - starti = job_index*step - endi = min((job_index+1)*step, all_line_number) +#XZ, 01/14/2009: This method is for parallel computing only. +#XZ: It is supposed to be called when "Genetic Correlation, Pearson's r" (method 1) +#XZ: or "Genetic Correlation, Spearman's rho" (method 2) is selected +def compute_corr( input_nnCorr, input_trait, input_list, computing_method): + + allcorrelations = [] + + for line in input_list: + tokens = line.split('","') + tokens[-1] = tokens[-1][:-2] #remove the last " + tokens[0] = tokens[0][1:] #remove the first " + + traitdataName = tokens[0] + database_trait = tokens[1:] + + if computing_method == "1": #XZ: Pearson's r + corr,nOverlap = utility.webqtlUtil.calCorrelationText(input_trait, database_trait, input_nnCorr) + else: #XZ: Spearman's rho + corr,nOverlap = utility.webqtlUtil.calCorrelationRankText(input_trait, database_trait, input_nnCorr) + traitinfo = [traitdataName,corr,nOverlap] + allcorrelations.append(traitinfo) + + return allcorrelations + +def get_correlation_method_key(form_data): + #XZ, 09/28/2008: if user select "1", then display 1, 3 and 4. + #XZ, 09/28/2008: if user select "2", then display 2, 3 and 5. + #XZ, 09/28/2008: if user select "3", then display 1, 3 and 4. + #XZ, 09/28/2008: if user select "4", then display 1, 3 and 4. + #XZ, 09/28/2008: if user select "5", then display 2, 3 and 5. + + method = form_data.formdata.getvalue("method") + if method not in ["1", "2", "3" ,"4", "5"]: + return "1" + + return method + + +def get_custom_trait(form_data, cursor): + """Pulls the custom trait, if it exists, out of the form data""" + trait_name = form_data.formdata.getvalue('fullname') + + if trait_name: + trait = webqtlTrait(fullname=trait_name, cursor=cursor) + trait.retrieveInfo() + return trait + else: + return None + + +#XZ, 09/18/2008: get the information such as value, variance of the input strain names from the form. +def get_sample_data(form_data): + if form_data.allstrainlist: + mdpchoice = form_data.formdata.getvalue('MDPChoice') + #XZ, in HTML source code, it is "BXD Only" or "BXH only", and so on + if mdpchoice == "1": + strainlist = form_data.f1list + form_data.strainlist + #XZ, in HTML source code, it is "MDP Only" + elif mdpchoice == "2": + strainlist = [] + strainlist2 = form_data.f1list + form_data.strainlist + for strain in form_data.allstrainlist: + if strain not in strainlist2: + strainlist.append(strain) + #So called MDP Panel + if strainlist: + strainlist = form_data.f1list+form_data.parlist+strainlist + #XZ, in HTML source code, it is "All Cases" + else: + strainlist = form_data.allstrainlist + #XZ, 09/18/2008: put the trait data into dictionary form_data.allTraitData + form_data.readData(form_data.allstrainlist) + else: + mdpchoice = None + strainlist = form_data.strainlist + #XZ, 09/18/2008: put the trait data into dictionary form_data.allTraitData + form_data.readData() + + return strainlist + + +def get_mdp_choice(form_data): + if form_data.allstrainlist: + return form_data.formdata.getvalue("MDPChoice") + else: + return None + + +def get_species(fd, cursor): + #XZ, 3/16/2010: variable RISet must be pass by the form + RISet = fd.RISet + #XZ, 12/12/2008: get species infomation + species = webqtlDatabaseFunction.retrieveSpecies(cursor=cursor, RISet=RISet) + return species + + +def sortTraitCorrelations(traits, method="1"): + for trait in traits: + if trait.lit_corr != None and float(trait.lit_corr) > 0.6: + debug_file.write(str(trait.lit_corr) + "\n") + if method in TISSUE_METHODS: + #traits.sort(key=lambda trait: trait.tissue_corr != None and abs(float(trait.tissue_corr)) or 0) + traits.sort(key=lambda trait: trait.tissue_corr != None and abs(trait.tissue_corr), reverse=True) + elif method == METHOD_LIT: + #traits.sort(key=lambda trait: trait.lit_corr != None and abs(float(trait.lit_corr)) or 0) + traits.sort(key=lambda trait: trait.lit_corr != None and abs(trait.lit_corr), reverse=True) + else: + #traits.sort(key=lambda trait: trait.correlation != None and abs(float(trait.correlation)) or 0) + traits.sort(key=lambda trait: trait.correlation != None and abs(trait.correlation), reverse=True) + + for trait in traits: + if trait.lit_corr != None and float(trait.lit_corr) > 0.6: + debug_file2.write(str(trait.lit_corr) + "\n") + + return traits + + + +def cmpTraitCorr(A,B): + try: + if abs(A.correlation) < abs(B.correlation): + return 1 + elif abs(A.correlation) == abs(B.correlation): + return 0 + else: + return -1 + except: + return 0 + +def auth_user_for_db(db, cursor, target_db_name, privilege, username): + """Authorize a user for access to a database if that database is + confidential. A db (identified by a record in ProbeSetFreeze) contains a + list of authorized users who may access it, as well as its confidentiality + level. + + If the current user's privilege level is greater than 'user', ie: root or + admin, then they are automatically authed, otherwise, check the + AuthorizedUsers field for the presence of their name.""" + + if db.type == 'ProbeSet': + cursor.execute('SELECT Id, Name, FullName, confidentiality, AuthorisedUsers FROM ProbeSetFreeze WHERE Name = "%s"' % target_db_name) + indId, indName, indFullName, confidential, AuthorisedUsers = cursor.fetchall()[0] + + if confidential: + authorized = 0 + + #for the dataset that confidentiality is 1 + #1. 'admin' and 'root' can see all of the dataset + #2. 'user' can see the dataset that AuthorisedUsers contains his id(stored in the Id field of User table) + if webqtlConfig.USERDICT[privilege] > webqtlConfig.USERDICT['user']: + authorized = 1 + else: + if username in AuthorisedUsers.split(","): + authorized = 1 - one_job_input_list = [] + if not authorized: + raise AuthException("The %s database you selected is not open to the public at this time, please go back and select other database." % indFullName) - for i in range( starti, endi ): - one_job_input_list.append( input_line_list[i] ) - job_input_lists.append( one_job_input_list ) +class CorrelationPage(templatePage): - ppservers = () - # Creates jobserver with automatically detected number of workers - job_server = pp.Server(ppservers=ppservers) + corrMinInformative = 4 - jobs = [] - results = [] + PAGE_HEADING = "Correlation Table" + CORRELATION_METHODS = {"1" : "Genetic Correlation (Pearson's r)", + "2" : "Genetic Correlation (Spearman's rho)", + "3" : "SGO Literature Correlation", + "4" : "Tissue Correlation (Pearson's r)", + "5" : "Tissue Correlation (Spearman's rho)"} - for one_job_input_list in job_input_lists: #pay attention to modules from outside - jobs.append( job_server.submit(func=compute_corr, args=(nnCorr, _newvals, one_job_input_list, self.method), depfuncs=(), modules=("utility.webqtlUtil",)) ) + RANK_ORDERS = {"1": 0, "2": 1, "3": 0, "4": 0, "5": 1} - for one_job in jobs: - one_result = one_job() - results.append( one_result ) - for one_result in results: - for one_traitinfo in one_result: - allcorrelations.append( one_traitinfo ) + def error(self, message, error="Error", heading = None): + heading = heading or self.PAGE_HEADING + return templatePage.error(heading = heading, detail = [message], error=error) - datasetFile.close() - totalTraits = len(allcorrelations) - #except: - # useFastMethod = False - # self.error(heading="No computation method",detail="Something is wrong within the try except block in CorrelationPage python module.",error="Computation Error") - # return + def __init__(self, fd): - #XZ, 01/08/2009: use the original method to retrieve from database and compute. - if not useFastMethod: + # Call the superclass constructor + templatePage.__init__(self, fd) - traitdatabase, dataStartPos = self.fetchAllDatabaseData(species=species, GeneId=input_trait_GeneId, GeneSymbol=input_trait_symbol, strains=_strains, db=self.db, method=self.method, returnNumber=self.returnNumber, tissueProbeSetFreezeId=TissueProbeSetFreezeId) + # Connect to the database + if not self.openMysql(): + return - totalTraits = len(traitdatabase) #XZ, 09/18/2008: total trait number + # Read the genotype from a file + if not fd.genotype: + fd.readGenotype() - for traitdata in traitdatabase: - traitdataName = traitdata[0] - traitvals = traitdata[dataStartPos:] - if self.method == "1" or self.method == "3" or self.method == "4": - corr,nOverlap = webqtlUtil.calCorrelation(traitvals,_vals,nnCorr) - else: - corr,nOverlap = webqtlUtil.calCorrelationRank(traitvals,_vals,nnCorr) + sample_list = get_sample_data(fd) + mdp_choice = get_mdp_choice(fd) # No idea what this is yet + self.species = get_species(fd, self.cursor) - traitinfo = [traitdataName,corr,nOverlap] + #XZ, 09/18/2008: get all information about the user selected database. + target_db_name = fd.formdata.getvalue('database') + self.target_db_name = target_db_name - #XZ, 09/28/2008: if user select '3', then fetchAllDatabaseData would give us LitCorr in the [1] position - #XZ, 09/28/2008: if user select '4' or '5', then fetchAllDatabaseData would give us Tissue Corr in the [1] position - #XZ, 09/28/2008: and Tissue Corr P Value in the [2] position - if input_trait_GeneId and self.db.type == "ProbeSet": - if self.method == "3": - traitinfo.append( traitdata[1] ) - if self.method == "4" or self.method == "5": - traitinfo.append( traitdata[1] ) - traitinfo.append( traitdata[2] ) + try: + self.db = webqtlDataset(target_db_name, self.cursor) + except: + detail = ["The database you just requested has not been established yet."] + self.error(detail) + return - allcorrelations.append(traitinfo) + # Auth if needed + try: + auth_user_for_db(self.db, self.cursor, target_db_name, self.privilege, self.userName) + except AuthException, e: + detail = [e.message] + return self.error(detail) + #XZ, 09/18/2008: filter out the strains that have no value. + self.sample_names, vals, vars, N = fd.informativeStrains(sample_list) -############################################################# + #CF - If less than a minimum number of strains/cases in common, don't calculate anything + if len(self.sample_names) < self.corrMinInformative: + detail = ['Fewer than %d strain data were entered for %s data set. No calculation of correlation has been attempted.' % (self.corrMinInformative, fd.RISet)] + self.error(heading=PAGE_HEADING,detail=detail) - if self.method == "3" and input_trait_GeneId: - allcorrelations.sort(webqtlUtil.cmpLitCorr) - #XZ, 3/31/2010: Theoretically, we should create one function 'comTissueCorr' - #to compare each trait by their tissue corr p values. - #But because the tissue corr p values are generated by permutation test, - #the top ones always have p value 0. So comparing p values actually does nothing. - #In addition, for the tissue data in our database, the N is always the same. - #So it's safe to compare with tissue corr statistic value. - #That's the same as literature corr. - elif self.method in ["4","5"] and input_trait_GeneId: - allcorrelations.sort(webqtlUtil.cmpLitCorr) - else: - allcorrelations.sort(webqtlUtil.cmpCorr) + self.method = get_correlation_method_key(fd) + correlation_method = self.CORRELATION_METHODS[self.method] + rankOrder = self.RANK_ORDERS[self.method] - #XZ, 09/20/2008: we only need the top ones. - self.returnNumber = min(self.returnNumber,len(allcorrelations)) - allcorrelations = allcorrelations[:self.returnNumber] + # CF - Number of results returned + self.returnNumber = int(fd.formdata.getvalue('criteria')) - addLiteratureCorr = False - addTissueCorr = False + self.record_count = 0 - traitList = [] - for item in allcorrelations: - thisTrait = webqtlTrait(db=self.db, name=item[0], cursor=self.cursor) - thisTrait.retrieveInfo( QTL='Yes' ) + myTrait = get_custom_trait(fd, self.cursor) - nOverlap = item[2] - corr = item[1] - #XZ: calculate corrPValue - if nOverlap < 3: - corrPValue = 1.0 - else: - if abs(corr) >= 1.0: - corrPValue = 0.0 - else: - ZValue = 0.5*log((1.0+corr)/(1.0-corr)) - ZValue = ZValue*sqrt(nOverlap-3) - corrPValue = 2.0*(1.0 - reaper.normp(abs(ZValue))) - - thisTrait.Name = item[0] - thisTrait.corr = corr - thisTrait.nOverlap = nOverlap - thisTrait.corrPValue = corrPValue - # NL, 07/19/2010 - # js function changed, add a new parameter rankOrder for js function 'showTissueCorrPlot' - rankOrder = 0; - if self.method in ["2","5"]: - rankOrder = 1; - thisTrait.rankOrder =rankOrder - - #XZ, 26/09/2008: Method is 4 or 5. Have fetched tissue corr, but no literature correlation yet. - if len(item) == 5: - thisTrait.tissueCorr = item[3] - thisTrait.tissuePValue = item[4] - addLiteratureCorr = True + # We will not get Literature Correlations if there is no GeneId because there is nothing to look against + self.gene_id = int(fd.formdata.getvalue('GeneId') or 0) - #XZ, 26/09/2008: Method is 3, Have fetched literature corr, but no tissue corr yet. - elif len(item) == 4: - thisTrait.LCorr = item[3] - thisTrait.mouse_geneid = self.translateToMouseGeneID(species, thisTrait.geneid) - addTissueCorr = True + # We will not get Tissue Correlations if there is no gene symbol because there is nothing to look against + self.trait_symbol = myTrait.symbol + - #XZ, 26/09/2008: Method is 1 or 2. Have NOT fetched literature corr and tissue corr yet. - # Phenotype data will not have geneid, and neither will some probes - # we need to handle this because we will get an attribute error - else: - if input_trait_mouse_geneid and self.db.type=="ProbeSet": - addLiteratureCorr = True - if input_trait_symbol and self.db.type=="ProbeSet": - addTissueCorr = True + #XZ, 12/12/2008: if the species is rat or human, translate the geneid to mouse geneid + self.input_trait_mouse_gene_id = self.translateToMouseGeneID(self.species, self.gene_id) - traitList.append(thisTrait) + #XZ: As of Nov/13/2010, this dataset is 'UTHSC Illumina V6.2 RankInv B6 D2 average CNS GI average (May 08)' + self.tissue_probeset_freeze_id = 1 - if addLiteratureCorr: - traitList = self.getLiteratureCorrelationByList(input_trait_mouse_geneid, species, traitList) - if addTissueCorr: - traitList = self.getTissueCorrelationByList( primaryTraitSymbol=input_trait_symbol, traitList=traitList,TissueProbeSetFreezeId =TissueProbeSetFreezeId, method=self.method) + traitList = self.correlate(vals) + + _log.info("Done doing correlation calculation") -######################################################## +############################################################################################################################################ TD_LR = HT.TD(height=200,width="100%",bgColor='#eeeeee') mainfmName = webqtlUtil.genRandStr("fm_") form = HT.Form(cgi= os.path.join(webqtlConfig.CGIDIR, webqtlConfig.SCRIPTFILE), enctype='multipart/form-data', name= mainfmName, submit=HT.Input(type='hidden')) - hddn = {'FormID':'showDatabase', 'ProbeSetID':'_','database':self.target_db_name, 'databaseFull':self.db.fullname, 'CellID':'_', 'RISet':RISet, 'identification':fd.identification} + hddn = {'FormID': 'showDatabase', + 'ProbeSetID': '_', + 'database': self.target_db_name, + 'databaseFull': self.db.fullname, + 'CellID': '_', + 'RISet': fd.RISet, + 'identification': fd.identification} if myTrait: hddn['fullname']=fd.formdata.getvalue('fullname') - if mdpchoice: - hddn['MDPChoice']=mdpchoice + if mdp_choice: + hddn['MDPChoice']=mdp_choice #XZ, 09/18/2008: pass the trait data to next page by hidden parameters. @@ -510,13 +415,13 @@ class CorrelationPage(templatePage): #XZ, 3/11/2010: add one parameter to record if the method is rank order. form.append(HT.Input(name="rankOrder", value="%s" % rankOrder, type='hidden')) - form.append(HT.Input(name="TissueProbeSetFreezeId", value="%s" % TissueProbeSetFreezeId, type='hidden')) + form.append(HT.Input(name="TissueProbeSetFreezeId", value="%s" % self.tissue_probeset_freeze_id, type='hidden')) #################################### # generate the info on top of page # #################################### - info = self.getTopInfo(myTrait=myTrait, method=self.method, db=self.db, target_db_name=self.target_db_name, returnNumber=self.returnNumber, methodDict=methodDict, totalTraits=totalTraits, identification=fd.identification ) + info = self.getTopInfo(myTrait=myTrait, method=self.method, db=self.db, target_db_name=self.target_db_name, returnNumber=self.returnNumber, methodDict=self.CORRELATION_METHODS, totalTraits=traitList, identification=fd.identification ) ############## # Excel file # @@ -536,7 +441,7 @@ class CorrelationPage(templatePage): ##################################################################### - + #Select All, Deselect All, Invert Selection, Add to Collection mintmap = HT.Href(url="#redirect", onClick="databaseFunc(document.getElementsByName('%s')[0], 'showIntMap');" % mainfmName) mintmap_img = HT.Image("/images/multiple_interval_mapping1_final.jpg", name='mintmap', alt="Multiple Interval Mapping", title="Multiple Interval Mapping", style="border:none;") mintmap.append(mintmap_img) @@ -555,10 +460,10 @@ class CorrelationPage(templatePage): partialCorr = HT.Href(url="#redirect", onClick="databaseFunc(document.getElementsByName('%s')[0], 'partialCorrInput');" % mainfmName) partialCorr_img = HT.Image("/images/partial_correlation_final.jpg", name='partialCorr', alt="Partial Correlation", title="Partial Correlation", style="border:none;") partialCorr.append(partialCorr_img) - addselect = HT.Href(url="#redirect", onClick="addRmvSelection('%s', document.getElementsByName('%s')[0], 'addToSelection');" % (RISet, mainfmName)) + addselect = HT.Href(url="#redirect", onClick="addRmvSelection('%s', document.getElementsByName('%s')[0], 'addToSelection');" % (fd.RISet, mainfmName)) addselect_img = HT.Image("/images/add_collection1_final.jpg", name="addselect", alt="Add To Collection", title="Add To Collection", style="border:none;") addselect.append(addselect_img) - selectall = HT.Href(url="#redirect", onClick="checkAll(document.getElementsByName('%s')[0]);" % mainfmName) + selectall = HT.Href(url="#redirect", onClick="checkAll(document.getElementsByName('%s')[0]);" % mainfmName) selectall_img = HT.Image("/images/select_all2_final.jpg", name="selectall", alt="Select All", title="Select All", style="border:none;") selectall.append(selectall_img) selectinvert = HT.Href(url="#redirect", onClick = "checkInvert(document.getElementsByName('%s')[0]);" % mainfmName) @@ -575,20 +480,90 @@ class CorrelationPage(templatePage): selectandor.append(('OR','or')) selectandor.selected.append('AND') - pageTable = HT.TableLite(cellSpacing=0,cellPadding=0,width="100%", border=0, align="Left") - containerTable = HT.TableLite(cellSpacing=0,cellPadding=0,width="90%",border=0, align="Left") + #External analysis tools + GCATButton = HT.Href(url="#redirect", onClick="databaseFunc(document.getElementsByName('%s')[0], 'GCAT');" % mainfmName) + GCATButton_img = HT.Image("/images/GCAT_logo_final.jpg", name="GCAT", alt="GCAT", title="GCAT", style="border:none") + GCATButton.append(GCATButton_img) + + ODE = HT.Href(url="#redirect", onClick="databaseFunc(document.getElementsByName('%s')[0], 'ODE');" % mainfmName) + ODE_img = HT.Image("/images/ODE_logo_final.jpg", name="ode", alt="ODE", title="ODE", style="border:none") + ODE.append(ODE_img) + + ''' + #XZ, 07/07/2010: I comment out this block of code. + WebGestaltScript = HT.Script(language="Javascript") + WebGestaltScript.append(""" +setTimeout('openWebGestalt()', 2000); +function openWebGestalt(){ +var thisForm = document['WebGestalt']; +makeWebGestaltTree(thisForm, '%s', %d, 'edag_only.php'); +} + """ % (mainfmName, len(traitList))) + ''' + + self.cursor.execute('SELECT GeneChip.GO_tree_value FROM GeneChip, ProbeFreeze, ProbeSetFreeze WHERE GeneChip.Id = ProbeFreeze.ChipId and ProbeSetFreeze.ProbeFreezeId = ProbeFreeze.Id and ProbeSetFreeze.Name = "%s"' % self.db.name) + result = self.cursor.fetchone() + + if result: + GO_tree_value = result[0] + + if GO_tree_value: + + WebGestalt = HT.Href(url="#redirect", onClick="databaseFunc(document.getElementsByName('%s')[0], 'GOTree');" % mainfmName) + WebGestalt_img = HT.Image("/images/webgestalt_icon_final.jpg", name="webgestalt", alt="Gene Set Analysis Toolkit", title="Gene Set Analysis Toolkit", style="border:none") + WebGestalt.append(WebGestalt_img) + + hddnWebGestalt = { + 'id_list':'', + 'correlation':'', + 'id_value':'', + 'llid_list':'', + 'id_type':GO_tree_value, + 'idtype':'', + 'species':'', + 'list':'', + 'client':''} + + hddnWebGestalt['ref_type'] = hddnWebGestalt['id_type'] + hddnWebGestalt['cat_type'] = 'GO' + hddnWebGestalt['significancelevel'] = 'Top10' + + if self.species == 'rat': + hddnWebGestalt['org'] = 'Rattus norvegicus' + elif self.species == 'human': + hddnWebGestalt['org'] = 'Homo sapiens' + elif self.species == 'mouse': + hddnWebGestalt['org'] = 'Mus musculus' + else: + hddnWebGestalt['org'] = '' + + for key in hddnWebGestalt.keys(): + form.append(HT.Input(name=key, value=hddnWebGestalt[key], type='hidden')) - optionsTable = HT.TableLite(cellSpacing=2, cellPadding=0,width="320", height="80", border=0, align="Left") - optionsTable.append(HT.TR(HT.TD(selectall), HT.TD(reset), HT.TD(selectinvert), HT.TD(addselect), align="left")) - optionsTable.append(HT.TR(HT.TD(" "*1,"Select"), HT.TD("Deselect"), HT.TD(" "*1,"Invert"), HT.TD(" "*3,"Add"))) + + #Create tables with options, etc + + pageTable = HT.TableLite(cellSpacing=0,cellPadding=0,width="100%", border=0, align="Left") + + containerTable = HT.TableLite(cellSpacing=0,cellPadding=0,width="90%",border=0, align="Left") + + + if not GO_tree_value: + optionsTable = HT.TableLite(cellSpacing=2, cellPadding=0,width="480", height="80", border=0, align="Left") + optionsTable.append(HT.TR(HT.TD(selectall), HT.TD(reset), HT.TD(selectinvert), HT.TD(addselect), HT.TD(GCATButton), HT.TD(ODE), align="left")) + optionsTable.append(HT.TR(HT.TD(" "*1,"Select"), HT.TD("Deselect"), HT.TD(" "*1,"Invert"), HT.TD(" "*3,"Add"), HT.TD("Gene Set"), HT.TD(" "*2,"GCAT"))) + else: + optionsTable = HT.TableLite(cellSpacing=2, cellPadding=0,width="560", height="80", border=0, align="Left") + optionsTable.append(HT.TR(HT.TD(selectall), HT.TD(reset), HT.TD(selectinvert), HT.TD(addselect), HT.TD(GCATButton), HT.TD(ODE), HT.TD(WebGestalt), align="left")) + optionsTable.append(HT.TR(HT.TD(" "*1,"Select"), HT.TD("Deselect"), HT.TD(" "*1,"Invert"), HT.TD(" "*3,"Add"), HT.TD("Gene Set"), HT.TD(" "*2,"GCAT"), HT.TD(" "*3, "ODE"))) containerTable.append(HT.TR(HT.TD(optionsTable))) functionTable = HT.TableLite(cellSpacing=2,cellPadding=0,width="480",height="80", border=0, align="Left") functionRow = HT.TR(HT.TD(networkGraph, width="16.7%"), HT.TD(cormatrix, width="16.7%"), HT.TD(partialCorr, width="16.7%"), HT.TD(mcorr, width="16.7%"), HT.TD(mintmap, width="16.7%"), HT.TD(heatmap), align="left") labelRow = HT.TR(HT.TD(" "*1,HT.Text("Graph")), HT.TD(" "*1,HT.Text("Matrix")), HT.TD(" "*1,HT.Text("Partial")), HT.TD(HT.Text("Compare")), HT.TD(HT.Text("QTL Map")), HT.TD(HT.Text(text="Heat Map"))) functionTable.append(functionRow, labelRow) - containerTable.append(HT.TR(HT.TD(functionTable), HT.BR())) + containerTable.append(HT.TR(HT.TD(functionTable), HT.BR())) #more_options = HT.Image("/images/more_options1_final.jpg", name='more_options', alt="Expand Options", title="Expand Options", style="border:none;", Class="toggleShowHide") @@ -597,12 +572,14 @@ class CorrelationPage(templatePage): moreOptions = HT.Input(type='button',name='options',value='More Options', onClick="",Class="toggle") fewerOptions = HT.Input(type='button',name='options',value='Fewer Options', onClick="",Class="toggle") + """ if (fd.formdata.getvalue('showHideOptions') == 'less'): - containerTable.append(HT.TR(HT.TD(" "), height="10"), HT.TR(HT.TD(HT.Div(fewerOptions, Class="toggleShowHide")))) - containerTable.append(HT.TR(HT.TD(" "))) + containerTable.append(HT.TR(HT.TD(" "), height="10"), HT.TR(HT.TD(HT.Div(fewerOptions, Class="toggleShowHide")))) + containerTable.append(HT.TR(HT.TD(" "))) else: - containerTable.append(HT.TR(HT.TD(" "), height="10"), HT.TR(HT.TD(HT.Div(moreOptions, Class="toggleShowHide")))) - containerTable.append(HT.TR(HT.TD(" "))) + containerTable.append(HT.TR(HT.TD(" "), height="10"), HT.TR(HT.TD(HT.Div(moreOptions, Class="toggleShowHide")))) + containerTable.append(HT.TR(HT.TD(" "))) + """ containerTable.append(HT.TR(HT.TD(HT.Span(selecttraits,' with r > ',selectgt, ' ',selectandor, ' r < ',selectlt,Class="bd1 cbddf fs11")), style="display:none;", Class="extra_options")) @@ -615,9 +592,9 @@ class CorrelationPage(templatePage): if self.db.type=="Geno": - containerTable.append(HT.TR(HT.TD(xlsUrl, height=40))) + containerTable.append(HT.TR(HT.TD(xlsUrl, height=40))) - pageTable.append(HT.TR(HT.TD(containerTable))) + pageTable.append(HT.TR(HT.TD(containerTable))) tblobj['header'], worksheet = self.getTableHeaderForGeno( method=self.method, worksheet=worksheet, newrow=newrow, headingStyle=headingStyle) newrow += 1 @@ -636,9 +613,9 @@ class CorrelationPage(templatePage): div = HT.Div(webqtlUtil.genTableObj(tblobj=tblobj, file=filename, sortby=sortby, tableID = "sortable", addIndex = "1"), corrScript, Id="sortable") - pageTable.append(HT.TR(HT.TD(div))) + pageTable.append(HT.TR(HT.TD(div))) - form.append(HT.Input(name='ShowStrains',type='hidden', value =1), + form.append(HT.Input(name='ShowStrains',type='hidden', value =1), HT.Input(name='ShowLine',type='hidden', value =1), HT.P(), HT.P(), pageTable) TD_LR.append(corrHeading, info, form, HT.P()) @@ -649,9 +626,9 @@ class CorrelationPage(templatePage): elif self.db.type=="Publish": - containerTable.append(HT.TR(HT.TD(xlsUrl, height=40))) + containerTable.append(HT.TR(HT.TD(xlsUrl, height=40))) - pageTable.append(HT.TR(HT.TD(containerTable))) + pageTable.append(HT.TR(HT.TD(containerTable))) tblobj['header'], worksheet = self.getTableHeaderForPublish(method=self.method, worksheet=worksheet, newrow=newrow, headingStyle=headingStyle) newrow += 1 @@ -661,7 +638,7 @@ class CorrelationPage(templatePage): corrScript = HT.Script(language="Javascript") corrScript.append("var corrArray = new Array();") - tblobj['body'], worksheet, corrScript = self.getTableBodyForPublish(traitList=traitList, formName=mainfmName, worksheet=worksheet, newrow=newrow, corrScript=corrScript, species=species) + tblobj['body'], worksheet, corrScript = self.getTableBodyForPublish(traitList=traitList, formName=mainfmName, worksheet=worksheet, newrow=newrow, corrScript=corrScript, species=self.species) workbook.close() @@ -671,7 +648,7 @@ class CorrelationPage(templatePage): # NL, 07/27/2010. genTableObj function has been moved from templatePage.py to webqtlUtil.py; div = HT.Div(webqtlUtil.genTableObj(tblobj=tblobj, file=filename, sortby=sortby, tableID = "sortable", addIndex = "1"), corrScript, Id="sortable") - pageTable.append(HT.TR(HT.TD(div))) + pageTable.append(HT.TR(HT.TD(div))) form.append( HT.Input(name='ShowStrains',type='hidden', value =1), @@ -685,7 +662,6 @@ class CorrelationPage(templatePage): elif self.db.type=="ProbeSet": - tblobj['header'], worksheet = self.getTableHeaderForProbeSet(method=self.method, worksheet=worksheet, newrow=newrow, headingStyle=headingStyle) newrow += 1 @@ -694,7 +670,7 @@ class CorrelationPage(templatePage): corrScript = HT.Script(language="Javascript") corrScript.append("var corrArray = new Array();") - tblobj['body'], worksheet, corrScript = self.getTableBodyForProbeSet(traitList=traitList, primaryTrait=myTrait, formName=mainfmName, worksheet=worksheet, newrow=newrow, corrScript=corrScript, species=species) + tblobj['body'], worksheet, corrScript = self.getTableBodyForProbeSet(traitList=traitList, primaryTrait=myTrait, formName=mainfmName, worksheet=worksheet, newrow=newrow, corrScript=corrScript, species=self.species) workbook.close() objfile = open('%s.obj' % (webqtlConfig.TMPDIR+filename), 'wb') @@ -741,95 +717,21 @@ class CorrelationPage(templatePage): #updated by NL. Delete function generateJavaScript, move js files to dhtml.js, webqtl.js and jqueryFunction.js #variables: filename, strainIds and vals are required by getquerystring function - strainIds=self.getStrainIds(species=species, strains=_strains) + strainIds=self.getStrainIds(species=self.species, strains=self.sample_names) var1 = HT.Input(name="filename", value=filename, type='hidden') var2 = HT.Input(name="strainIds", value=strainIds, type='hidden') - var3 = HT.Input(name="vals", value=_vals, type='hidden') + var3 = HT.Input(name="vals", value=vals, type='hidden') customizerButton = HT.Input(type="button", Class="button", value="Add Correlation", onClick = "xmlhttpPost('%smain.py?FormID=AJAX_table', 'sortable', (getquerystring(this.form)))" % webqtlConfig.CGIDIR) containerTable.append(HT.TR(HT.TD(HT.Span(var1,var2,var3,customizerButton, "with", dbCustomizer, Class="bd1 cbddf fs11"), HT.BR(), HT.BR()), style="display:none;", Class="extra_options")) - #outside analysis part - GCATButton = HT.Href(url="#redirect", onClick="databaseFunc(document.getElementsByName('%s')[0], 'GCAT');" % mainfmName) - GCATButton_img = HT.Image("/images/GCAT_logo_final.jpg", name="GCAT", alt="GCAT", title="GCAT", style="border:none") - GCATButton.append(GCATButton_img) + containerTable.append(HT.TR(HT.TD(xlsUrl, HT.BR(), HT.BR()))) - ODE = HT.Href(url="#redirect", onClick="databaseFunc(document.getElementsByName('%s')[0], 'ODE');" % mainfmName) - ODE_img = HT.Image("/images/ODE_logo_final.jpg", name="ode", alt="ODE", title="ODE", style="border:none") - ODE.append(ODE_img) + pageTable.append(HT.TR(HT.TD(containerTable))) - ''' - #XZ, 07/07/2010: I comment out this block of code. - WebGestaltScript = HT.Script(language="Javascript") - WebGestaltScript.append(""" -setTimeout('openWebGestalt()', 2000); -function openWebGestalt(){ - var thisForm = document['WebGestalt']; - makeWebGestaltTree(thisForm, '%s', %d, 'edag_only.php'); -} - """ % (mainfmName, len(traitList))) - ''' - - self.cursor.execute('SELECT GeneChip.GO_tree_value FROM GeneChip, ProbeFreeze, ProbeSetFreeze WHERE GeneChip.Id = ProbeFreeze.ChipId and ProbeSetFreeze.ProbeFreezeId = ProbeFreeze.Id and ProbeSetFreeze.Name = "%s"' % self.db.name) - result = self.cursor.fetchone() + pageTable.append(HT.TR(HT.TD(div))) - if result: - GO_tree_value = result[0] - - if GO_tree_value: - - WebGestalt = HT.Href(url="#redirect", onClick="databaseFunc(document.getElementsByName('%s')[0], 'GOTree');" % mainfmName) - WebGestalt_img = HT.Image("/images/webgestalt_icon_final.jpg", name="webgestalt", alt="Gene Set Analysis Toolkit", title="Gene Set Analysis Toolkit", style="border:none") - WebGestalt.append(WebGestalt_img) - - hddnWebGestalt = { - 'id_list':'', - 'correlation':'', - 'id_value':'', - 'llid_list':'', - 'id_type':GO_tree_value, - 'idtype':'', - 'species':'', - 'list':'', - 'client':''} - - hddnWebGestalt['ref_type'] = hddnWebGestalt['id_type'] - hddnWebGestalt['cat_type'] = 'GO' - hddnWebGestalt['significancelevel'] = 'Top10' - - if species == 'rat': - hddnWebGestalt['org'] = 'Rattus norvegicus' - elif species == 'human': - hddnWebGestalt['org'] = 'Homo sapiens' - elif species == 'mouse': - hddnWebGestalt['org'] = 'Mus musculus' - else: - hddnWebGestalt['org'] = '' - - for key in hddnWebGestalt.keys(): - form.append(HT.Input(name=key, value=hddnWebGestalt[key], type='hidden')) - - - LinkOutTable = HT.TableLite(cellSpacing=2,cellPadding=0,width="320",height="80", border=0, align="Left") - - if not GO_tree_value: - LinkOutRow = HT.TR(HT.TD(GCATButton, width="50%"), HT.TD(ODE, width="50%"), align="left") - LinkOutLabels = HT.TR(HT.TD(" ", HT.Text("GCAT"), width="50%"), HT.TD(" ",HT.Text("ODE"), width="50%"), align="left") - else: - LinkOutRow = HT.TR(HT.TD(WebGestalt, width="25%"), HT.TD(GCATButton, width="25%"), HT.TD(ODE, width="25%"), align="left") - LinkOutLabels = HT.TR(HT.TD(HT.Text("Gene Set")), HT.TD(" "*2, HT.Text("GCAT")), HT.TD(" "*3, HT.Text("ODE")), style="display:none;", Class="extra_options") - - LinkOutTable.append(LinkOutRow,LinkOutLabels) - - containerTable.append(HT.TR(HT.TD(LinkOutTable), Class="extra_options", style="display:none;")) - - containerTable.append(HT.TR(HT.TD(xlsUrl, HT.BR(), HT.BR()))) - - pageTable.append(HT.TR(HT.TD(containerTable))) - - pageTable.append(HT.TR(HT.TD(div))) - - if species == 'human': + if self.species == 'human': heatmap = "" form.append(HT.Input(name='ShowStrains',type='hidden', value =1), @@ -1009,7 +911,7 @@ Resorting this table <br> tempTable = self.getTempLiteratureTable(species=species, input_species_geneid=GeneId, returnNumber=returnNumber) if method == "4" or method == "5": - tempTable = self.getTempTissueCorrTable(primaryTraitSymbol=GeneSymbol, TissueProbeSetFreezeId=tissueProbeSetFreezeId, method=method, returnNumber=returnNumber) + tempTable = self.getTempTissueCorrTable(primaryTraitSymbol=GeneSymbol, TissueProbeSetFreezeId=TISSUE_MOUSE_DB, method=method, returnNumber=returnNumber) for step in range(nnn): temp = [] @@ -1068,18 +970,30 @@ Resorting this table <br> oridata.append(results) datasize = len(oridata[0]) - traitdatabase = [] - # put all of the seperate data together into a huge list of lists + traits = [] + # put all of the separate data together into a huge list of lists for j in range(datasize): traitdata = list(oridata[0][j]) for i in range(1,nnn): traitdata += list(oridata[i][j][dataStartPos:]) - traitdatabase.append(traitdata) + + trait = Trait(traitdata[0], traitdata[dataStartPos:]) + + if method == METHOD_LIT: + trait.lit_corr = traitdata[1] + + if method in TISSUE_METHODS: + trait.tissue_corr = traitdata[1] + trait.p_tissue = traitdata[2] + + traits.append(trait) if tempTable: self.cursor.execute( 'DROP TEMPORARY TABLE %s' % tempTable ) - return traitdatabase, dataStartPos + return traits + + # XZ, 09/20/2008: This function creates TEMPORARY TABLE tmpTableName_2 and return its name. @@ -1159,7 +1073,7 @@ Resorting this table <br> except: return 0 - symbolCorrDict, symbolPvalueDict = self.calculateCorrOfAllTissueTrait(primaryTraitSymbol=primaryTraitSymbol, TissueProbeSetFreezeId=TissueProbeSetFreezeId, method=method) + symbolCorrDict, symbolPvalueDict = self.calculateCorrOfAllTissueTrait(primaryTraitSymbol=primaryTraitSymbol, TissueProbeSetFreezeId=TISSUE_MOUSE_DB, method=method) symbolCorrList = symbolCorrDict.items() @@ -1216,7 +1130,7 @@ Resorting this table <br> Returns a dictionary of 'TraitID':(tissueCorr, tissuePValue) for the requested correlation""" - tempTable = self.getTempTissueCorrTable(primaryTraitSymbol=primaryTraitSymbol, TissueProbeSetFreezeId=TissueProbeSetFreezeId, method=method, returnNumber=returnNumber) + tempTable = self.getTempTissueCorrTable(primaryTraitSymbol=primaryTraitSymbol, TissueProbeSetFreezeId=TISSUE_MOUSE_DB, method=method, returnNumber=returnNumber) query = "SELECT ProbeSet.Name, %s.Correlation, %s.PValue" % (tempTable, tempTable) query += ' FROM (ProbeSet, ProbeSetXRef, ProbeSetFreeze)' @@ -1276,6 +1190,225 @@ Resorting this table <br> return traitList + def get_trait(self, cached, vals): + + if cached: + _log.info("Using the fast method because the file exists") + lit_corrs = {} + tissue_corrs = {} + use_lit = False + if self.method == METHOD_LIT: + lit_corrs = self.fetchLitCorrelations(species=self.species, GeneId=self.gene_id, db=self.db, returnNumber=self.returnNumber) + use_lit = True + + use_tissue_corr = False + if self.method in TISSUE_METHODS: + tissue_corrs = self.fetchTissueCorrelations(db=self.db, primaryTraitSymbol=self.trait_symbol, TissueProbeSetFreezeId=TISSUE_MOUSE_DB, method=self.method, returnNumber = self.returnNumber) + use_tissue_corr = True + + DatabaseFileName = self.getFileName( target_db_name=self.target_db_name ) + datasetFile = open(webqtlConfig.TEXTDIR+DatabaseFileName,'r') + + #XZ, 01/08/2009: read the first line + line = datasetFile.readline() + cached_sample_names = webqtlUtil.readLineCSV(line)[1:] + + #XZ, 01/08/2009: This step is critical. It is necessary for this new method. + #XZ: The original function fetchAllDatabaseData uses all strains stored in variable _strains to + #XZ: retrieve the values of each strain from database in real time. + #XZ: The new method uses all strains stored in variable dataset_strains to create a new variable + #XZ: _newvals. _newvals has the same length as dataset_strains. The items in _newvals is in + #XZ: the same order of items in dataset_strains. The value of each item in _newvals is either + #XZ: the value of correspinding strain in _vals or 'None'. + new_vals = [] + for name in cached_sample_names: + if name in self.sample_names: + new_vals.append(float(vals[self.sample_names.index(name)])) + else: + new_vals.append('None') + + nnCorr = len(new_vals) + + #XZ, 01/14/2009: If literature corr or tissue corr is selected, + #XZ: there is no need to use parallel computing. + + traits = [] + data_start = 1 + for line in datasetFile: + raw_trait = webqtlUtil.readLineCSV(line) + trait = Trait.from_csv(raw_trait, data_start) + trait.lit_corr = lit_corrs.get(trait.name) + trait.tissue_corr, trait.p_tissue = tissue_corrs.get(trait.name, (None, None)) + traits.append(trait) + + return traits, new_vals + + else: + _log.info("Using the slow method for correlation") + + _log.info("Fetching from database") + traits = self.fetchAllDatabaseData(species=self.species, GeneId=self.gene_id, GeneSymbol=self.trait_symbol, strains=self.sample_names, db=self.db, method=self.method, returnNumber=self.returnNumber, tissueProbeSetFreezeId= self.tissue_probeset_freeze_id) + _log.info("Done fetching from database") + totalTraits = len(traits) #XZ, 09/18/2008: total trait number + + return traits, vals + + + def do_parallel_correlation(self): + _log.info("Invoking parallel computing") + input_line_list = datasetFile.readlines() + _log.info("Read lines from the file") + all_line_number = len(input_line_list) + + step = 1000 + job_number = math.ceil( float(all_line_number)/step ) + + job_input_lists = [] + + _log.info("Configuring jobs") + + for job_index in range( int(job_number) ): + starti = job_index*step + endi = min((job_index+1)*step, all_line_number) + + one_job_input_list = [] + + for i in range( starti, endi ): + one_job_input_list.append( input_line_list[i] ) + + job_input_lists.append( one_job_input_list ) + + _log.info("Creating pp servers") + + ppservers = () + # Creates jobserver with automatically detected number of workers + job_server = pp.Server(ppservers=ppservers) + + _log.info("Done creating servers") + + jobs = [] + results = [] + + _log.info("Starting parallel computation, submitting jobs") + for one_job_input_list in job_input_lists: #pay attention to modules from outside + jobs.append( job_server.submit(func=compute_corr, args=(nnCorr, _newvals, one_job_input_list, self.method), depfuncs=(), modules=("utility.webqtlUtil",)) ) + _log.info("Done submitting jobs") + + for one_job in jobs: + one_result = one_job() + results.append( one_result ) + + _log.info("Acquiring results") + + for one_result in results: + for one_traitinfo in one_result: + allcorrelations.append( one_traitinfo ) + + _log.info("Appending the results") + + datasetFile.close() + totalTraits = len(allcorrelations) + _log.info("Done correlating using the fast method") + + + def correlate(self, vals): + + correlations = [] + + #XZ: Use the fast method only for probeset dataset, and this dataset must have been created. + #XZ: Otherwise, use original method + _log.info("Entering correlation") + + db_filename = self.getFileName( target_db_name=self.target_db_name ) + + cache_available = db_filename in os.listdir(webqtlConfig.TEXTDIR) + + # If the cache file exists, do a cached correlation for probeset data + if self.db.type == "ProbeSet": +# if self.method in [METHOD_SAMPLE_PEARSON, METHOD_SAMPLE_RANK] and cache_available: +# traits = do_parallel_correlation() +# +# else: + + (traits, vals) = self.get_trait(cache_available, vals) + + for trait in traits: + trait.calculate_correlation(vals, self.method) + + self.record_count = len(traits) #ZS: This isn't a good way to get this value, so I need to change it later + + #XZ, 3/31/2010: Theoretically, we should create one function 'comTissueCorr' + #to compare each trait by their tissue corr p values. + #But because the tissue corr p values are generated by permutation test, + #the top ones always have p value 0. So comparing p values actually does nothing. + #In addition, for the tissue data in our database, the N is always the same. + #So it's safe to compare with tissue corr statistic value. + #That's the same as literature corr. + #if self.method in [METHOD_LIT, METHOD_TISSUE_PEARSON, METHOD_TISSUE_RANK] and self.gene_id: + # traits.sort(webqtlUtil.cmpLitCorr) + #else: + #if self.method in TISSUE_METHODS: + # sort(traits, key=lambda A: math.fabs(A.tissue_corr)) + #elif self.method == METHOD_LIT: + # traits.sort(traits, key=lambda A: math.fabs(A.lit_corr)) + #else: + traits = sortTraitCorrelations(traits, self.method) + + # Strip to the top N correlations + traits = traits[:min(self.returnNumber, len(traits))] + + addLiteratureCorr = False + addTissueCorr = False + + trait_list = [] + for trait in traits: + db_trait = webqtlTrait(db=self.db, name=trait.name, cursor=self.cursor) + db_trait.retrieveInfo( QTL='Yes' ) + + db_trait.Name = trait.name + db_trait.corr = trait.correlation + db_trait.nOverlap = trait.overlap + db_trait.corrPValue = trait.p_value + + # NL, 07/19/2010 + # js function changed, add a new parameter rankOrder for js function 'showTissueCorrPlot' + db_trait.RANK_ORDER = self.RANK_ORDERS[self.method] + + #XZ, 26/09/2008: Method is 4 or 5. Have fetched tissue corr, but no literature correlation yet. + if self.method in TISSUE_METHODS: + db_trait.tissueCorr = trait.tissue_corr + db_trait.tissuePValue = trait.p_tissue + addTissueCorr = True + + + #XZ, 26/09/2008: Method is 3, Have fetched literature corr, but no tissue corr yet. + elif self.method == METHOD_LIT: + db_trait.LCorr = trait.lit_corr + db_trait.mouse_geneid = self.translateToMouseGeneID(self.species, db_trait.geneid) + addLiteratureCorr = True + + #XZ, 26/09/2008: Method is 1 or 2. Have NOT fetched literature corr and tissue corr yet. + # Phenotype data will not have geneid, and neither will some probes + # we need to handle this because we will get an attribute error + else: + if self.input_trait_mouse_gene_id and self.db.type=="ProbeSet": + addLiteratureCorr = True + if self.trait_symbol and self.db.type=="ProbeSet": + addTissueCorr = True + + trait_list.append(db_trait) + + if addLiteratureCorr: + trait_list = self.getLiteratureCorrelationByList(self.input_trait_mouse_gene_id, + self.species, trait_list) + if addTissueCorr: + trait_list = self.getTissueCorrelationByList( + primaryTraitSymbol = self.trait_symbol, + traitList = trait_list, + TissueProbeSetFreezeId = TISSUE_MOUSE_DB, + method=self.method) + + return trait_list def calculateCorrOfAllTissueTrait(self, primaryTraitSymbol=None, TissueProbeSetFreezeId=None, method=None): @@ -1283,10 +1416,10 @@ Resorting this table <br> symbolCorrDict = {} symbolPvalueDict = {} - primaryTraitSymbolValueDict = correlationFunction.getGeneSymbolTissueValueDictForTrait(cursor=self.cursor, GeneNameLst=[primaryTraitSymbol], TissueProbeSetFreezeId=TissueProbeSetFreezeId) + primaryTraitSymbolValueDict = correlationFunction.getGeneSymbolTissueValueDictForTrait(cursor=self.cursor, GeneNameLst=[primaryTraitSymbol], TissueProbeSetFreezeId=TISSUE_MOUSE_DB) primaryTraitValue = primaryTraitSymbolValueDict.values()[0] - SymbolValueDict = correlationFunction.getGeneSymbolTissueValueDictForTrait(cursor=self.cursor, GeneNameLst=[], TissueProbeSetFreezeId=TissueProbeSetFreezeId) + SymbolValueDict = correlationFunction.getGeneSymbolTissueValueDictForTrait(cursor=self.cursor, GeneNameLst=[], TissueProbeSetFreezeId=TISSUE_MOUSE_DB) if method in ["2","5"]: symbolCorrDict, symbolPvalueDict = correlationFunction.batchCalTissueCorr(primaryTraitValue,SymbolValueDict,method='spearman') @@ -1301,7 +1434,7 @@ Resorting this table <br> #XZ, 10/13/2010 def getTissueCorrelationByList(self, primaryTraitSymbol=None, traitList=None, TissueProbeSetFreezeId=None, method=None): - primaryTraitSymbolValueDict = correlationFunction.getGeneSymbolTissueValueDictForTrait(cursor=self.cursor, GeneNameLst=[primaryTraitSymbol], TissueProbeSetFreezeId=TissueProbeSetFreezeId) + primaryTraitSymbolValueDict = correlationFunction.getGeneSymbolTissueValueDictForTrait(cursor=self.cursor, GeneNameLst=[primaryTraitSymbol], TissueProbeSetFreezeId=TISSUE_MOUSE_DB) if primaryTraitSymbol.lower() in primaryTraitSymbolValueDict: primaryTraitValue = primaryTraitSymbolValueDict[primaryTraitSymbol.lower()] @@ -1312,7 +1445,7 @@ Resorting this table <br> if hasattr(thisTrait, 'symbol'): geneSymbolList.append(thisTrait.symbol) - SymbolValueDict = correlationFunction.getGeneSymbolTissueValueDictForTrait(cursor=self.cursor, GeneNameLst=geneSymbolList, TissueProbeSetFreezeId=TissueProbeSetFreezeId) + SymbolValueDict = correlationFunction.getGeneSymbolTissueValueDictForTrait(cursor=self.cursor, GeneNameLst=geneSymbolList, TissueProbeSetFreezeId=TISSUE_MOUSE_DB) for thisTrait in traitList: if hasattr(thisTrait, 'symbol') and thisTrait.symbol and thisTrait.symbol.lower() in SymbolValueDict: @@ -1339,7 +1472,7 @@ Resorting this table <br> if myTrait: if method in ["1","2"]: #genetic correlation info = HT.Paragraph("Values of Record %s in the " % myTrait.getGivenName(), HT.Href(text=myTrait.db.fullname,url=webqtlConfig.INFOPAGEHREF % myTrait.db.name,target="_blank", Class="fwn"), - " database were compared to all %d records in the " % totalTraits, HT.Href(text=db.fullname,url=webqtlConfig.INFOPAGEHREF % target_db_name,target="_blank", Class="fwn"), + " database were compared to all %d records in the " % self.record_count, HT.Href(text=db.fullname,url=webqtlConfig.INFOPAGEHREF % target_db_name,target="_blank", Class="fwn"), ' database. The top %d correlations ranked by the %s are displayed.' % (returnNumber,methodDict[method]), ' You can resort this list using the small arrowheads in the top row.') else: @@ -1359,7 +1492,7 @@ Resorting this table <br> ' You can resort this list using the small arrowheads in the top row.' ) elif identification: - info = HT.Paragraph('Values of %s were compared to all %d traits in ' % (identification, totalTraits), + info = HT.Paragraph('Values of %s were compared to all %d traits in ' % (identification, self.record_count), HT.Href(text=db.fullname,url=webqtlConfig.INFOPAGEHREF % target_db_name,target="_blank",Class="fwn"), ' database. The TOP %d correlations ranked by the %s are displayed.' % (returnNumber,methodDict[method]), ' You can resort this list using the small arrowheads in the top row.') @@ -1934,7 +2067,7 @@ Resorting this table <br> TCorr = thisTrait.tissueCorr TCorrStr = "%2.3f" % thisTrait.tissueCorr # NL, 07/19/2010: add a new parameter rankOrder for js function 'showTissueCorrPlot' - rankOrder = thisTrait.rankOrder + rankOrder = self.RANK_ORDERS[self.method] TCorrPlotURL = "javascript:showTissueCorrPlot('%s','%s','%s',%d)" %(formName, primaryTrait.symbol, thisTrait.symbol,rankOrder) tr.append(TDCell(HT.TD(HT.Href(text=TCorrStr, url=TCorrPlotURL, Class="fs12 fwn ff1"), Class="fs12 fwn ff1 b1 c222", align='right'), TCorrStr, abs(TCorr))) else: |