def importBioMarkerGeneAssociations():
try:
biomarker_files = unique.read_directory('BuildDBs/BioMarkers/')
except Exception:
biomarker_files = unique.read_directory('/BuildDBs/BioMarkers/')
x=0; marker_symbol_db={}
for file in biomarker_files:
if '.txt' in file:
fn = filepath('BuildDBs/BioMarkers/'+file)
for line in open(fn,'rU').xreadlines():
data = cleanUpLine(line)
t = string.split(data,'\t')
if x==0:
x = 1; y=0
for i in t:
if 'marker-in' in i: mi = y
if 'Symbol' in i: sy = y
y+=1
ensembl = t[0]; symbol = string.lower(t[sy]); marker = t[mi]
markers = string.split(marker,'|')
for marker in markers:
try: marker_symbol_db[marker].append(symbol)
except Exception: marker_symbol_db[marker]=[symbol]
marker_symbol_db = gene_associations.eliminate_redundant_dict_values(marker_symbol_db)
return marker_symbol_db
def importPAZARcompiled():
""" Skips over the above function when these tf-target file is downlaoded directly """
tf_dir = 'BuildDBs/PAZAR/symbol/tf-target.txt'
tf_to_symbol={}
fn = filepath(tf_dir)
for line in open(fn,'rU').xreadlines():
data = cleanUpLine(line)
tf_id,symbol = string.split(data,'\t')
try: tf_to_symbol[tf_id].append(symbol)
except Exception: tf_to_symbol[tf_id] = [symbol]
tf_to_symbol = gene_associations.eliminate_redundant_dict_values(tf_to_symbol)
return tf_to_symbol
def importAmandeusPredictions(force):
if force == 'yes':
downloadAmadeusPredictions()
x=0
tf_symbol_db={}
fn = filepath('BuildDBs/Amadeus/symbol-Metazoan-Amadeus.txt')
for line in open(fn,'rU').xreadlines():
data = cleanUpLine(line)
if x==0: x=1
else:
symbol,system,tf_name = string.split(data,'\t')
symbol = string.lower(symbol)
try: tf_symbol_db[tf_name].append(symbol)
except Exception: tf_symbol_db[tf_name]=[symbol]
tf_symbol_db = gene_associations.eliminate_redundant_dict_values(tf_symbol_db)
return tf_symbol_db
def importPhenotypeOntologyGeneAssociations():
x=0
pheno_symbol={}; phen=[]
fn = filepath('BuildDBs/Pheno/HMD_HumanPhenotype.rpt')
for line in open(fn,'rU').xreadlines():
data = cleanUpLine(line)
if x==0: x=1
else:
t = string.split(data,'\t')
hs_symbol=t[0]; hs_entrez=t[1]; mm_symbol=t[2]; mgi=t[3]; pheno_ids=t[4]
hs_symbol = string.lower(hs_symbol)
mm_symbol = string.lower(mm_symbol)
symbols = [mm_symbol,hs_symbol]
pheno_ids = string.split(pheno_ids,' '); phen+=pheno_ids
for pheno_id in pheno_ids:
if len(pheno_id)>0:
for symbol in symbols:
try: pheno_symbol[pheno_id].append(symbol)
except Exception: pheno_symbol[pheno_id]=[symbol]
phen = unique.unique(phen)
pheno_symbol = gene_associations.eliminate_redundant_dict_values(pheno_symbol)
return pheno_symbol
def importPAZARAssociations():
pazar_files = unique.read_directory('/BuildDBs/PAZAR')
species_db={}
tf_to_target={}
for file in pazar_files:
if '.csv' in file:
name = string.join(string.split(file,'_')[1:-1],'_')
fn = filepath('BuildDBs/PAZAR/'+file)
for line in open(fn,'rU').xreadlines():
data = cleanUpLine(line)
try:
### Each line contains the following 11 tab-delim fields:
### Fields are: <PAZAR TF ID> <TF Name> <PAZAR Gene ID> <ensembl gene accession> <chromosome> <gene start coordinate> <gene end coordinate> <species> <project name> <PMID> <analysis method>
pazar_tf_id, tf_name, pazar_geneid, ens_gene, chr, gene_start,gene_end,species,project,pmid,analysis_method = string.split(data,'\t')
species,genus = string.split(species,' ')
species = species[0]+genus[0]
tft=TFTargetInfo(tf_name,ens_gene,project,pmid,analysis_method)
try: tf_to_target[species,tf_name].append(tft)
except Exception: tf_to_target[species,tf_name] = [tft]
species_db[species]=[]
except Exception:
None ### Occurs due to file formatting issues (during an update?)
determine_tf_geneids = 'no'
if determine_tf_geneids == 'yes':
""" The below code is probably most useful for creation of complex regulatory inference networks in Cytoscape """
uniprot_ensembl_db = importUniProtAnnotations(species_db)
missing=[]
tf_to_target_ens={}
for (species,tf_name) in tf_to_target:
original_tf_name = tf_name
try:
ens_gene = uniprot_ensembl_db[species,tf_name]
tf_to_target_ens[ens_gene]=tf_to_target[species,tf_name]
except Exception:
try:
tf_name = string.split(tf_name,'_')[0]
ens_gene = uniprot_ensembl_db[species,tf_name]
tf_to_target_ens[ens_gene]=tf_to_target[species,original_tf_name]
except Exception:
try:
tf_names=[]
if '/' in tf_name:
tf_names = string.split(tf_name,'/')
elif ' ' in tf_name:
tf_names = string.split(tf_name,' ')
for tf_name in tf_names:
ens_gene = uniprot_ensembl_db[species,tf_name]
tf_to_target_ens[ens_gene]=tf_to_target[species,original_tf_name]
except Exception: missing.append((tf_name,species))
print 'Ensembl IDs found for UniProt Transcription factor names:',len(tf_to_target_ens),'and missing:', len(missing)
#print missing[:20]
### Translate all species data to gene symbol to export for all species
species_tf_targets={}
for (species,tf_name) in tf_to_target:
try:
tf_db = species_tf_targets[species]
tf_db[tf_name] = tf_to_target[species,tf_name]
except Exception:
tf_db = {}
tf_db[tf_name] = tf_to_target[species,tf_name]
species_tf_targets[species] = tf_db
tf_dir = 'BuildDBs/PAZAR/symbol/tf-target.txt'
tf_data = export.ExportFile(tf_dir)
tf_to_symbol={}
#print 'Exporting:',tf_dir
#print len(species_tf_targets)
for species in species_tf_targets:
try: gene_to_source_id = gene_associations.getGeneToUid(species,('hide','Ensembl-Symbol'))
except Exception: gene_to_source_id={}
tf_db = species_tf_targets[species]
for tf_name in tf_db:
for tft in tf_db[tf_name]:
try:
for symbol in gene_to_source_id[tft.Ensembl()]:
symbol = string.lower(symbol)
tf_id = tf_name+'(Source:'+tft.Project()+'-PAZAR'+')'
tf_data.write(tf_id+'\t'+symbol+'\n')
try: tf_to_symbol[tf_id].append(symbol)
except Exception: tf_to_symbol[tf_id] = [symbol]
except Exception: null=[];
tf_data.close()
tf_to_symbol = gene_associations.eliminate_redundant_dict_values(tf_to_symbol)
return tf_to_symbol