### Based on code from AltAnalyze's LineageProfiler (http://altanalyze.org)
#Author Nathan Salomonis - nsalomonis@gmail.com
#Permission is hereby granted, free of charge, to any person obtaining a copy
#of this software and associated documentation files (the "Software"), to deal
#in the Software without restriction, including without limitation the rights
#to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
#copies of the Software, and to permit persons to whom the Software is furnished
#to do so, subject to the following conditions:
#THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED,
#INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A
#PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT
#HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION
#OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
#SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
"""
This script iterates the LineageProfiler algorithm (correlation based classification method) to identify sample types relative to one
of two references given one or more gene models. The main function is runLineageProfiler.
The program performs the following actions:
1) Import a tab-delimited reference expression file with three columns (ID, biological group 1, group 2) and a header row (biological group names)
2) Import a tab-delimited expression file with gene IDs (column 1), sample names (row 1) and normalized expression values (e.g., delta CT values)
3) (optional - import existing models) Import a tab-delimited file with comma delimited gene-models for analysis
4) (optional - find new models) Identify all possible combinations of gene models for a supplied model size variable (e.g., --s 7)
5) Iterate through any supplied or identified gene models to obtain predictions for novel or known sample types
6) Export prediction results for all analyzed models to the folder SampleClassification.
7) (optional) Print the top 20 scores and models for all possible model combinations of size --s
"""
import sys, string
import math
import os.path
import copy
import time
import getopt
import scipy
try: import unique ### Not required (used in AltAnalyze)
except Exception: None
try: import export ### Not required (used in AltAnalyze)
except Exception: None
#import salstat_stats; reload(salstat_stats)
try:
from scipy import stats
use_scipy = True
except Exception:
use_scipy = False ### scipy is not required but is used as a faster implementation of Fisher Exact Test when present
def filepath(filename):
try: fn = unique.filepath(filename)
except Exception: fn = filename
return fn
def exportFile(filename):
try: export_data = export.ExportFile(filename)
except Exception: export_data = open(filename,'w')
return export_data
def makeUnique(item):
db1={}; list1=[]; k=0
for i in item:
try: db1[i]=[]
except TypeError: db1[tuple(i)]=[]; k=1
for i in db1:
if k==0: list1.append(i)
else: list1.append(list(i))
list1.sort()
return list1
def cleanUpLine(line):
line = string.replace(line,'\n','')
line = string.replace(line,'\c','')
data = string.replace(line,'\r','')
data = string.replace(data,'"','')
return data
def returnLargeGlobalVars():
### Prints all large global variables retained in memory (taking up space)
all = [var for var in globals() if (var[:2], var[-2:]) != ("__", "__")]
for var in all:
try:
if len(globals()[var])>1:
print var, len(globals()[var])
except Exception: null=[]
def clearObjectsFromMemory(db_to_clear):
db_keys={}
for key in db_to_clear: db_keys[key]=[]
for key in db_keys:
try: del db_to_clear[key]
except Exception:
try:
for i in key: del i ### For lists of tuples
except Exception: del key ### For plain lists
def int_check(value):
val_float = float(value)
val_int = int(value)
if val_float == val_int:
integer_check = 'yes'
if val_float != val_int:
integer_check = 'no'
return integer_check
def IQR(array):
k1 = 75
k2 = 25
array.sort()
n = len(array)
value1 = float((n*k1)/100)
value2 = float((n*k2)/100)
if int_check(value1) == 'no':
k1_val = int(value1) + 1
if int_check(value1) == 'yes':
k1_val = int(value1)
if int_check(value2) == 'no':
k2_val = int(value2) + 1
if int_check(value2) == 'yes':
k2_val = int(value2)
median_val = scipy.median(array)
upper75th = array[k1_val]
lower25th = array[k2_val]
int_qrt_range = upper75th - lower25th
T1 = lower25th-(1.5*int_qrt_range)
T2 = upper75th+(1.5*int_qrt_range)
return lower25th,median_val,upper75th,int_qrt_range,T1,T2
class IQRData:
def __init__(self,maxz,minz,medz,iq1,iq3):
self.maxz = maxz; self.minz = minz
self.medz = medz; self.iq1 = iq1
self.iq3 = iq3
def Max(self): return self.maxz
def Min(self): return self.minz
def Medium(self): return self.medz
def IQ1(self): return self.iq1
def IQ3(self): return self.iq3
def SummaryValues(self):
vals = string.join([str(self.IQ1()),str(self.Min()),str(self.Medium()),str(self.Max()),str(self.IQ3())],'\t')
return vals
def importGeneModels(geneModels):
fn=filepath(geneModels); x=0
geneModels=[]
for line in open(fn,'rU').xreadlines():
genes = cleanUpLine(line)
genes = string.replace(genes,"'",'')
genes = string.replace(genes,' ',',')
genes = string.split(genes,',')
models=[]
for gene in genes:
if len(gene)>0:
models.append(gene)
geneModels.append(models)
return geneModels
######### Below code deals is specific to this module #########
def runLineageProfiler(species,array_type,exp_input,exp_output,codingtype,compendium_platform,modelSize=None,customMarkers=False,geneModels=False):
""" This code differs from LineageProfiler.py in that it is able to iterate through the LineageProfiler functions with distinct geneModels
that are either supplied by the user or discovered from all possible combinations. """
global exp_output_file; exp_output_file = exp_output; global targetPlatform
global tissues; global sample_headers
global analysis_type; global coding_type; coding_type = codingtype
global tissue_to_gene; tissue_to_gene = {}; global platform; global cutoff
global customMarkerFile; global delim; global keyed_by
#global tissue_specific_db
customMarkerFile = customMarkers
if geneModels == False: geneModels = []
else:
geneModels = importGeneModels(geneModels)
if '\\' in exp_input: delim = '\\'
elif '//' in exp_input: delim = '//'
else: delim = "/"
print '\nRunning LineageProfiler analysis on',string.split(exp_input,delim)[-1][:-4]
global correlate_by_order; correlate_by_order = 'no'
global rho_threshold; rho_threshold = -1
global correlate_to_tissue_specific; correlate_to_tissue_specific = 'no'
platform = array_type
cutoff = 0.01
global value_type
if 'stats.' in exp_input:
value_type = 'calls'
else:
value_type = 'expression'
tissue_specific_db={}; sample_headers=[]; tissues=[]
if len(array_type)==2:
### When a user-supplied expression is provided (no ExpressionOutput files provided - importGeneIDTranslations)
vendor, array_type = array_type
platform = array_type
else: vendor = 'Not needed'
if 'RawSplice' in exp_input or 'FullDatasets' in exp_input or coding_type == 'AltExon':
analysis_type = 'AltExon'
if platform != compendium_platform: ### If the input IDs are not Affymetrix Exon 1.0 ST probesets, then translate to the appropriate system
translate_to_genearray = 'no'
targetPlatform = compendium_platform
translation_db = importExonIDTranslations(array_type,species,translate_to_genearray)
keyed_by = 'translation'
else: translation_db=[]; keyed_by = 'primaryID'; targetPlatform = compendium_platform
elif array_type == "3'array" or array_type == 'AltMouse':
### Get arrayID to Ensembl associations
if vendor != 'Not needed':
### When no ExpressionOutput files provided (user supplied matrix)
translation_db = importVendorToEnsemblTranslations(species,vendor,exp_input)
else:
translation_db = importGeneIDTranslations(exp_output)
keyed_by = 'translation'
targetPlatform = compendium_platform
analysis_type = 'geneLevel'
else:
translation_db=[]; keyed_by = 'primaryID'; targetPlatform = compendium_platform; analysis_type = 'geneLevel'
targetPlatform = compendium_platform ### Overides above
try: importTissueSpecificProfiles(species,tissue_specific_db)
except Exception:
try:
try:
targetPlatform = 'exon'
importTissueSpecificProfiles(species,tissue_specific_db)
except Exception:
try:
targetPlatform = 'gene'
importTissueSpecificProfiles(species,tissue_specific_db)
except Exception:
targetPlatform = "3'array"
importTissueSpecificProfiles(species,tissue_specific_db)
except ZeroDivisionError:
print 'No compatible compendiums present...'
print e
forceError
all_marker_genes=[]
for gene in tissue_specific_db:
all_marker_genes.append(gene)
if len(geneModels)>0:
allPossibleClassifiers = geneModels
elif modelSize == None or modelSize == 'optimize':
allPossibleClassifiers = [all_marker_genes]
else:
### A specific model size has been specified (e.g., find all 10-gene models)
allPossibleClassifiers = getRandomSets(all_marker_genes,modelSize)
num=1
if len(allPossibleClassifiers)<16:
print 'Using:'
for model in allPossibleClassifiers:
print 'model',num,model
num+=1
### This is the main analysis function
print 'Number of references to compare to:',len(tissues), tissues
if modelSize != 'optimize':
hit_list, hits, fails, prognostic_class_db,sample_diff_z, evaluate_size = interateLineageProfiler(exp_input, tissue_specific_db, allPossibleClassifiers,translation_db,modelSize)
else:
summary_hit_list=[]
evaluate_size = len(allPossibleClassifiers[0])
hit_list, hits, fails, prognostic_class_db,sample_diff_z, evaluate_size = interateLineageProfiler(exp_input, tissue_specific_db, allPossibleClassifiers,translation_db,None)
while evaluate_size > 4:
hit_list.sort()
top_model = hit_list[-1][-1]
top_model_score = hit_list[-1][0]
"""
try: ### Used for evaluation only - gives the same top models
second_model = hit_list[-2][-1]
second_model_score = hit_list[-2][0]
if second_model_score == top_model_score:
top_model = second_model_score ### Try this
print 'selecting secondary'
except Exception: None
"""
allPossibleClassifiers = [hit_list[-1][-1]]
hit_list, hits, fails, prognostic_class_db,sample_diff_z, evaluate_size = interateLineageProfiler(exp_input, tissue_specific_db, allPossibleClassifiers,translation_db,modelSize)
summary_hit_list+=hit_list
hit_list = summary_hit_list
root_dir = string.join(string.split(exp_output_file,'/')[:-1],'/')+'/'
dataset_name = string.replace(string.split(exp_input,'/')[-1][:-4],'exp.','')
output_classification_file = root_dir+'SampleClassification/'+dataset_name+'-SampleClassification.txt'
try: os.mkdir(root_dir+'SampleClassification')
except Exception: None
export_summary = exportFile(output_classification_file)
models = []
for i in allPossibleClassifiers:
i = string.replace(str(i),"'",'')[1:-1]
models.append(i)
headers = string.join(['Samples']+tissues+['Overall Prognostic Score','Median Z-score Difference']+models,'\t')+'\n'
export_summary.write(headers)
for sample in prognostic_class_db:
class_db = prognostic_class_db[sample]
class_scores=[]; class_scores_str=[]
for tissue in tissues:
class_scores_str.append(str(class_db[tissue]))
class_scores.append(class_db[tissue])
zscore_distribution = map(str,sample_diff_z[sample])
dist_list = map(float,zscore_distribution) ### convert to list
median_score = scipy.median(dist_list)
if len(tissues)==2:
class_scores_str = str(class1_score-class2_score) ### range of positive and negative scores for a two-class test
values = [sample]+class_scores_str+[str(max(class_scores)-min(class_scores)),str(median_score)]
#print values
#print zscore_distribution
export_summary.write(string.join(values+zscore_distribution,'\t')+'\n')
export_summary.close()
print 'Results file written to:',root_dir+'SampleClassification/'+dataset_name+'-SampleClassification.txt','\n'
hit_list.sort(); hit_list.reverse()
top_hit_list=[]
top_hit_db={}
hits_db={}; fails_db={}
for i in sample_diff_z:
zscore_distribution = sample_diff_z[i]
maxz = max(zscore_distribution); minz = min(zscore_distribution)
lower25th,medz,upper75th,int_qrt_range,T1,T2 = IQR(zscore_distribution)
if float(maxz)>float(T2): maxz = T2
if float(minz) < float(T1): minz = T1
iqr = IQRData(maxz,minz,medz,lower25th,upper75th)
#sample_diff_z[i] = iqr
sample_diff_z[i] = string.join(map(str,zscore_distribution),'\t')
for i in hits:
try: hits_db[i]+=1
except Exception: hits_db[i]=1
for i in fails:
try: fails_db[i]+=1
except Exception: fails_db[i]=1
for i in fails_db:
if i not in hits:
try:
#print i+'\t'+'0\t'+str(fails_db[i])+'\t'+ sample_diff_z[i]
None
except Exception:
#print i
None
if modelSize != False:
hits=[]
for i in hits_db:
hits.append([hits_db[i],i])
hits.sort()
hits.reverse()
for i in hits:
if i[1] in fails_db: fail = fails_db[i[1]]
else: fail = 0
try:
#print i[1]+'\t'+str(i[0])+'\t'+str(fail)+'\t'+sample_diff_z[i[1]]
None
except Exception:
#print i[1]
None
for i in hit_list:
if i[0]>0:
top_hit_list.append(i[-1])
top_hit_db[tuple(i[-1])]=i[0]
print 'Top 20 top hits'
for i in hit_list[:20]:
print i[:5],i[-1]
return top_hit_db
def interateLineageProfiler(exp_input,tissue_specific_db,allPossibleClassifiers,translation_db,modelSize):
hit_list=[]
### Iterate through LineageProfiler for all gene models (allPossibleClassifiers)
times = 1; k=1000; l=1000; hits=[]; fails=[]; f=0; s=0; sample_diff_z={}; prognostic_class1_db={}; prognostic_class2_db={}
prognostic_class_db={}
begin_time = time.time()
evaluate_size=len(allPossibleClassifiers[0]) ### Number of reference markers to evaluate
if modelSize=='optimize':
evaluate_size -= 1
allPossibleClassifiers = getRandomSets(allPossibleClassifiers[0],evaluate_size)
for classifiers in allPossibleClassifiers:
tissue_to_gene={}; expession_subset=[]; sample_headers=[]; classifier_specific_db={}
for gene in classifiers:
try: classifier_specific_db[gene] = tissue_specific_db[gene]
except Exception: None
expession_subset = importGeneExpressionValues(exp_input,classifier_specific_db,translation_db,expession_subset)
### If the incorrect gene system was indicated re-run with generic parameters
if len(expession_subset)==0:
translation_db=[]; keyed_by = 'primaryID'; targetPlatform = compendium_platform; analysis_type = 'geneLevel'
tissue_specific_db={}
importTissueSpecificProfiles(species,tissue_specific_db)
expession_subset = importGeneExpressionValues(exp_input,tissue_specific_db,translation_db,expession_subset)
if len(expession_subset)!=len(classifiers): f+=1
#if modelSize=='optimize': print len(expession_subset), len(classifiers);sys.exit()
if len(expession_subset)==len(classifiers): ### Sometimes a gene or two are missing from one set
s+=1
zscore_output_dir,tissue_scores = analyzeTissueSpecificExpressionPatterns(tissue_specific_db,expession_subset)
#except Exception: print len(classifier_specific_db), classifiers; error
headers = list(tissue_scores['headers']); del tissue_scores['headers']
if times == k:
end_time = time.time()
print int(end_time-begin_time),'seconds'
k+=l
times+=1; index=0; positive=0; positive_score_diff=0
sample_number = (len(headers)-1)
population1_denom=0; population1_pos=0; population2_pos=0; population2_denom=0
diff_positive=[]; diff_negative=[]
while index < sample_number:
scores = map(lambda x: tissue_scores[x][index], tissue_scores)
scores_copy = list(scores); scores_copy.sort()
diff_z = scores_copy[-1]-scores_copy[-2] ### Diff between the top two scores
j=0
for tissue in tissue_scores:
if scores[j] == max(scores):
hit_score = 1
else: hit_score = 0
if len(tissues)>2:
if tissue+':' in headers[index+1] and hit_score==1:
positive+=1
try:
class_db = prognostic_class_db[headers[index+1]]
try: class_db[tissue]+=hit_score
except Exception: class_db[tissue]=hit_score
except Exception:
class_db={}
class_db[tissue]=hit_score
prognostic_class_db[headers[index+1]] = class_db
j+=1
#diff_z = tissue_scores[tissues[0]][index]-tissue_scores[tissues[-1]][index]
try: sample_diff_z[headers[index+1]].append(diff_z)
except Exception: sample_diff_z[headers[index+1]] =[diff_z]
if len(tissues)==2:
if headers[index+1] not in prognostic_class1_db:
prognostic_class1_db[headers[index+1]]=0 ### Create a default value for each sample
if headers[index+1] not in prognostic_class2_db:
prognostic_class2_db[headers[index+1]]=0 ### Create a default value for each sample
if diff_z>0:
prognostic_class1_db[headers[index+1]]+=1
if diff_z<0:
prognostic_class2_db[headers[index+1]]+=1
if diff_z>0 and (tissues[0]+'-' in headers[index+1] or tissues[0]+':' in headers[index+1]):
positive+=1; positive_score_diff+=abs(diff_z)
population1_pos+=1; diff_positive.append(abs(diff_z))
hits.append(headers[index+1]) ### see which are correctly classified
elif diff_z<0 and (tissues[-1]+'-' in headers[index+1] or tissues[-1]+':' in headers[index+1]):
positive+=1; positive_score_diff+=abs(diff_z)
population2_pos+=1; diff_positive.append(abs(diff_z))
hits.append(headers[index+1]) ### see which are correctly classified
elif diff_z>0 and (tissues[-1]+'-' in headers[index+1] or tissues[-1]+':' in headers[index+1]): ### Incorrectly classified
diff_negative.append(abs(diff_z))
fails.append(headers[index+1])
elif diff_z<0 and (tissues[0]+'-' in headers[index+1] or tissues[0]+':' in headers[index+1]): ### Incorrectly classified
#print headers[index+1]
diff_negative.append(abs(diff_z))
fails.append(headers[index+1])
if (tissues[0]+'-' in headers[index+1] or tissues[0]+':' in headers[index+1]):
population1_denom+=1
else:
population2_denom+=1
index+=1
percent_positive = (float(positive)/float(index))*100
if len(tissues)==2:
hit_list.append([percent_positive,population1_pos, population1_denom,population2_pos,population2_denom,[avg(diff_positive),avg(diff_negative)],positive_score_diff,classifiers])
else:
hit_list.append([percent_positive,len(classifiers),classifiers])
return hit_list, hits, fails, prognostic_class_db, sample_diff_z, evaluate_size
def factorial(n):
### Code from http://docs.python.org/lib/module-doctest.html
if not n >= 0:
raise ValueError("n must be >= 0")
if math.floor(n) != n:
raise ValueError("n must be exact integer")
if n+1 == n: # catch a value like 1e300
raise OverflowError("n too large")
result = 1
factor = 2
while factor <= n:
result *= factor
factor += 1
return result
def choose(n,x):
"""Equation represents the number of ways in which x objects can be selected from a total of n objects without regard to order."""
#(n x) = n!/(x!(n-x)!)
f = factorial
result = f(n)/(f(x)*f(n-x))
return result
def getRandomSets(a,size):
#a = ['a','b','c','d','e','f','g','h','i','j','k','l','m','n','o','p','q','r','s','t','u','v','w','x','y','z']
#size = 4
select_set={'ENSG00000140678':'ITGAX','ENSG00000105835':'NAMPT','ENSG00000027697':'IFNGR1','ENSG00000120129':'DUSP1','ENSG00000003402':'CFLAR','ENSG00000113269':'RNF130'}
select_set={}
select_set2={'ENSG00000163602': 'RYBP'}
negative_select = {'ENSG00000105352':'CEACAM4'}
negative_select={}
import random
possible_sets = choose(len(a),size)
print 'Possible',size,'gene combinations to test',possible_sets
permute_ls = []; done = 0; permute_db={}
while done == 0:
b = list(tuple(a)); random.shuffle(b)
bx_set={}
i = 0
while i < len(b):
try:
bx = b[i:i+size]; bx.sort()
if len(bx)==size: permute_db[tuple(bx)]=None
else: break
except Exception: break
i+=1
if len(permute_db) == possible_sets:
done=1; break
for i in permute_db:
add=0; required=0; exclude=0
for l in i:
if len(select_set)>0:
if l in select_set: add+=1
#if l in select_set2: required+=1
#if l in negative_select: exclude+=1
else: add = 1000
if add>2 and exclude==0:# and required==1:
permute_ls.append(i)
#print len(permute_ls)
return permute_ls
def importVendorToEnsemblTranslations(species,vendor,exp_input):
translation_db={}
"""
### Faster method but possibly not as good
uid_db = simpleUIDImport(exp_input)
import gene_associations
### Use the same annotation method that is used to create the ExpressionOutput annotations
array_to_ens = gene_associations.filterGeneToUID(species,'Ensembl',vendor,associated_IDs)
for arrayid in array_to_ens:
ensembl_list = array_to_ens[arrayid]
try: translation_db[arrayid] = ensembl_list[0] ### This first Ensembl is ranked as the most likely valid based on various metrics in getArrayAnnotationsFromGOElite
except Exception: None
"""
translation_db={}
import BuildAffymetrixAssociations
### Use the same annotation method that is used to create the ExpressionOutput annotations
use_go = 'yes'
conventional_array_db={}
conventional_array_db = BuildAffymetrixAssociations.getArrayAnnotationsFromGOElite(conventional_array_db,species,vendor,use_go)
for arrayid in conventional_array_db:
ca = conventional_array_db[arrayid]
ens = ca.Ensembl()
try: translation_db[arrayid] = ens[0] ### This first Ensembl is ranked as the most likely valid based on various metrics in getArrayAnnotationsFromGOElite
except Exception: None
return translation_db
def importTissueSpecificProfiles(species,tissue_specific_db):
if analysis_type == 'AltExon':
filename = 'AltDatabase/ensembl/'+species+'/'+species+'_'+targetPlatform +'_tissue-specific_AltExon_protein_coding.txt'
else:
filename = 'AltDatabase/ensembl/'+species+'/'+species+'_'+targetPlatform +'_tissue-specific_'+coding_type+'.txt'
if customMarkerFile != False:
filename = customMarkerFile
if value_type == 'calls':
filename = string.replace(filename,'.txt','_stats.txt')
fn=filepath(filename); x=0
tissues_added={}
for line in open(fn,'rU').xreadlines():
data = cleanUpLine(line)
t = string.split(data,'\t')
if x==0:
print 'Importing the tissue compedium database:',string.split(filename,delim)[-1][:-4]
headers = t; x=1; index=0
for i in headers:
if 'UID' == i: ens_index = index; uid_index = index
if analysis_type == 'AltExon': ens_index = ens_index ### Assigned above when analyzing probesets
elif 'Ensembl' in i: ens_index = index
if 'marker-in' in i: tissue_index = index+1; marker_in = index
index+=1
try:
for i in t[tissue_index:]: tissues.append(i)
except Exception:
for i in t[1:]: tissues.append(i)
if keyed_by == 'primaryID':
try: ens_index = uid_index
except Exception: None
else:
try:
gene = t[0]
tissue_exp = map(float, t[1:])
tissue_specific_db[gene]=x,tissue_exp ### Use this to only grab relevant gene expression profiles from the input dataset
except Exception:
gene = string.split(t[ens_index],'|')[0] ### Only consider the first listed gene - this gene is the best option based on ExpressionBuilder rankings
#if 'Pluripotent Stem Cells' in t[marker_in] or 'Heart' in t[marker_in]:
#if t[marker_in] not in tissues_added: ### Only add the first instance of a gene for that tissue - used more for testing to quickly run the analysis
tissue_exp = map(float, t[tissue_index:])
if value_type == 'calls':
tissue_exp = produceDetectionCalls(tissue_exp,platform) ### 0 or 1 calls
tissue_specific_db[gene]=x,tissue_exp ### Use this to only grab relevant gene expression profiles from the input dataset
tissues_added[t[marker_in]]=[]
x+=1
print len(tissue_specific_db), 'genes in the tissue compendium database'
if correlate_to_tissue_specific == 'yes':
try: importTissueCorrelations(filename)
except Exception:
null=[]
#print '\nNo tissue-specific correlations file present. Skipping analysis.'; kill
return tissue_specific_db
def importTissueCorrelations(filename):
filename = string.replace(filename,'specific','specific_correlations')
fn=filepath(filename); x=0
for line in open(fn,'rU').xreadlines():
data = cleanUpLine(line)
if x==0: x=1 ### Ignore header line
else:
uid,symbol,rho,tissue = string.split(data,'\t')
if float(rho)>rho_threshold: ### Variable used for testing different thresholds internally
try: tissue_to_gene[tissue].append(uid)
except Exception: tissue_to_gene[tissue] = [uid]
def simpleUIDImport(filename):
"""Import the UIDs in the gene expression file"""
uid_db={}
fn=filepath(filename)
for line in open(fn,'rU').xreadlines():
data = cleanUpLine(line)
uid_db[string.split(data,'\t')[0]]=[]
return uid_db
def importGeneExpressionValues(filename,tissue_specific_db,translation_db,expession_subset):
### Import gene-level expression raw values
fn=filepath(filename); x=0; genes_added={}; gene_expression_db={}
dataset_name = string.split(filename,delim)[-1][:-4]
#print 'importing:',dataset_name
for line in open(fn,'rU').xreadlines():
data = cleanUpLine(line)
t = string.split(data,'\t')
if x==0:
if '#' not in data:
for i in t[1:]: sample_headers.append(i)
x=1
else:
gene = t[0]
#if '-' not in gene and ':E' in gene: print gene;sys.exit()
if analysis_type == 'AltExon':
try: ens_gene,exon = string.split(gene,'-')[:2]
except Exception: exon = gene
gene = exon
if keyed_by == 'translation': ### alternative value is 'primaryID'
"""if gene == 'ENSMUSG00000025915-E19.3':
for i in translation_db: print [i], len(translation_db); break
print gene, [translation_db[gene]];sys.exit()"""
try: gene = translation_db[gene] ### Ensembl annotations
except Exception: gene = 'null'
if gene in tissue_specific_db:
index,tissue_exp=tissue_specific_db[gene]
try: genes_added[gene]+=1
except Exception: genes_added[gene]=1
try: exp_vals = map(float, t[1:])
except Exception:
### If a non-numeric value in the list
exp_vals=[]
for i in t[1:]:
try: exp_vals.append(float(i))
except Exception: exp_vals.append(i)
if value_type == 'calls': ### Hence, this is a DABG or RNA-Seq expression
exp_vals = produceDetectionCalls(exp_vals,targetPlatform) ### 0 or 1 calls
gene_expression_db[gene] = [index,exp_vals]
#print len(gene_expression_db), 'matching genes in the dataset and tissue compendium database'
for gene in genes_added:
if genes_added[gene]>1: del gene_expression_db[gene] ### delete entries that are present in the input set multiple times (not trustworthy)
else: expession_subset.append(gene_expression_db[gene]) ### These contain the rank order and expression
#print len(expession_subset);sys.exit()
expession_subset.sort() ### This order now matches that of
gene_expression_db=[]
return expession_subset
def produceDetectionCalls(values,Platform):
# Platform can be the compendium platform (targetPlatform) or analyzed data platform (platform or array_type)
new=[]
for value in values:
if Platform == 'RNASeq':
if value>1:
new.append(1) ### expressed
else:
new.append(0)
else:
if value<cutoff: new.append(1)
else: new.append(0)
return new
def importGeneIDTranslations(filename):
### Import ExpressionOutput/DATASET file to obtain Ensembl associations (typically for Affymetrix 3' arrays)
fn=filepath(filename); x=0; translation_db={}
for line in open(fn,'rU').xreadlines():
data = cleanUpLine(line)
t = string.split(data,'\t')
if x==0:
headers = t; x=1; index=0
for i in headers:
if 'Ensembl' in i: ens_index = index; break
index+=1
else:
uid = t[0]
ens_geneids = t[ens_index]
ens_geneid = string.split(ens_geneids,'|')[0] ### In v.2.0.5, the first ID is the best protein coding candidate
if len(ens_geneid)>0:
translation_db[uid] = ens_geneid
return translation_db
def remoteImportExonIDTranslations(array_type,species,translate_to_genearray,targetplatform):
global targetPlatform; targetPlatform = targetplatform
translation_db = importExonIDTranslations(array_type,species,translate_to_genearray)
return translation_db
def importExonIDTranslations(array_type,species,translate_to_genearray):
gene_translation_db={}; gene_translation_db2={}
if targetPlatform == 'gene' and translate_to_genearray == 'no':
### Get gene array to exon array probeset associations
gene_translation_db = importExonIDTranslations('gene',species,'yes')
for geneid in gene_translation_db:
exonid = gene_translation_db[geneid]
gene_translation_db2[exonid] = geneid
#print exonid, geneid
translation_db = gene_translation_db2
else:
filename = 'AltDatabase/'+species+'/'+array_type+'/'+species+'_'+array_type+'-exon_probesets.txt'
### Import exon array to target platform translations (built for DomainGraph visualization)
fn=filepath(filename); x=0; translation_db={}
print 'Importing the translation file',string.split(fn,delim)[-1][:-4]
for line in open(fn,'rU').xreadlines():
data = cleanUpLine(line)
t = string.split(data,'\t')
if x==0: x=1
else:
platform_id,exon_id = t
if targetPlatform == 'gene' and translate_to_genearray == 'no':
try:
translation_db[platform_id] = gene_translation_db[exon_id] ### return RNA-Seq to gene array probeset ID
#print platform_id, exon_id, gene_translation_db[exon_id];sys.exit()
except Exception: null=[]
else:
translation_db[platform_id] = exon_id
del gene_translation_db; del gene_translation_db2
return translation_db
def analyzeTissueSpecificExpressionPatterns(tissue_specific_db,expession_subset):
tissue_specific_sorted = []; genes_present={}; tissue_exp_db={}; gene_order_db={}; gene_order=[]
for (index,vals) in expession_subset: genes_present[index]=[]
for gene in tissue_specific_db:
tissue_specific_sorted.append(tissue_specific_db[gene])
gene_order_db[tissue_specific_db[gene][0]] = gene ### index order (this index was created before filtering)
tissue_specific_sorted.sort()
new_index=0
for (index,tissue_exp) in tissue_specific_sorted:
try:
null=genes_present[index]
i=0
gene_order.append([new_index,gene_order_db[index]]); new_index+=1
for f in tissue_exp:
### The order of the tissue specific expression profiles is based on the import gene order
try: tissue_exp_db[tissues[i]].append(f)
except Exception: tissue_exp_db[tissues[i]] = [f]
i+=1
except Exception: null=[] ### Gene is not present in the input dataset
### Organize sample expression, with the same gene order as the tissue expression set
sample_exp_db={}
for (index,exp_vals) in expession_subset:
i=0
for f in exp_vals:
### The order of the tissue specific expression profiles is based on the import gene order
try: sample_exp_db[sample_headers[i]].append(f)
except Exception: sample_exp_db[sample_headers[i]] = [f]
i+=1
if correlate_by_order == 'yes':
### Rather than correlate to the absolute expression order, correlate to the order of expression (lowest to highest)
sample_exp_db = replaceExpressionWithOrder(sample_exp_db)
tissue_exp_db = replaceExpressionWithOrder(tissue_exp_db)
global tissue_comparison_scores; tissue_comparison_scores={}
if correlate_to_tissue_specific == 'yes':
### Create a gene_index that reflects the current position of each gene
gene_index={}
for (i,gene) in gene_order: gene_index[gene] = i
### Create a tissue to gene-index from the gene_index
tissue_to_index={}
for tissue in tissue_to_gene:
for gene in tissue_to_gene[tissue]:
if gene in gene_index: ### Some are not in both tissue and sample datasets
index = gene_index[gene] ### Store by index, since the tissue and expression lists are sorted by index
try: tissue_to_index[tissue].append(index)
except Exception: tissue_to_index[tissue] = [index]
tissue_to_index[tissue].sort()
sample_exp_db,tissue_exp_db = returnTissueSpecificExpressionProfiles(sample_exp_db,tissue_exp_db,tissue_to_index)
PearsonCorrelationAnalysis(sample_exp_db,tissue_exp_db)
sample_exp_db=[]; tissue_exp_db=[]
zscore_output_dir,tissue_scores = exportCorrelationResults()
return zscore_output_dir, tissue_scores
def returnTissueSpecificExpressionProfiles(sample_exp_db,tissue_exp_db,tissue_to_index):
tissue_exp_db_abreviated={}
sample_exp_db_abreviated={} ### This db is designed differently than the non-tissue specific (keyed by known tissues)
### Build the tissue specific expression profiles
for tissue in tissue_exp_db:
tissue_exp_db_abreviated[tissue] = []
for index in tissue_to_index[tissue]:
tissue_exp_db_abreviated[tissue].append(tissue_exp_db[tissue][index]) ### populate with just marker expression profiles
### Build the sample specific expression profiles
for sample in sample_exp_db:
sample_tissue_exp_db={}
sample_exp_db[sample]
for tissue in tissue_to_index:
sample_tissue_exp_db[tissue] = []
for index in tissue_to_index[tissue]:
sample_tissue_exp_db[tissue].append(sample_exp_db[sample][index])
sample_exp_db_abreviated[sample] = sample_tissue_exp_db
return sample_exp_db_abreviated, tissue_exp_db_abreviated
def replaceExpressionWithOrder(sample_exp_db):
for sample in sample_exp_db:
sample_exp_sorted=[]; i=0
for exp_val in sample_exp_db[sample]: sample_exp_sorted.append([exp_val,i]); i+=1
sample_exp_sorted.sort(); sample_exp_resort = []; order = 0
for (exp_val,i) in sample_exp_sorted: sample_exp_resort.append([i,order]); order+=1
sample_exp_resort.sort(); sample_exp_sorted=[] ### Order lowest expression to highest
for (i,o) in sample_exp_resort: sample_exp_sorted.append(o) ### The expression order replaces the expression, in the original order
sample_exp_db[sample] = sample_exp_sorted ### Replace exp with order
return sample_exp_db
def PearsonCorrelationAnalysis(sample_exp_db,tissue_exp_db):
#print "Beginning LineageProfiler analysis"
k=0
original_increment = int(len(tissue_exp_db)/15.00); increment = original_increment
p = 1 ### Default value if not calculated
for tissue in tissue_exp_db:
#print k,"of",len(tissue_exp_db),"classifier tissue/cell-types"
if k == increment: increment+=original_increment; #print '*',
k+=1
tissue_expression_list = tissue_exp_db[tissue]
for sample in sample_exp_db:
if correlate_to_tissue_specific == 'yes':
### Keyed by tissue specific sample profiles
sample_expression_list = sample_exp_db[sample][tissue] ### dictionary as the value for sample_exp_db[sample]
#print tissue, sample_expression_list
#print tissue_expression_list; sys.exit()
else: sample_expression_list = sample_exp_db[sample]
try:
### p-value is likely useful to report (not supreemly accurate but likely sufficient)
rho,p = stats.pearsonr(tissue_expression_list,sample_expression_list)
try: tissue_comparison_scores[tissue].append([rho,p,sample])
except Exception: tissue_comparison_scores[tissue] = [[rho,p,sample]]
except Exception:
### simple pure python implementation - no scipy required (not as fast though and no p-value)
#rho = pearson(tissue_expression_list,sample_expression_list)
None
#tst = salstat_stats.TwoSampleTests(tissue_expression_list,sample_expression_list)
#pp,pr = tst.PearsonsCorrelation()
#sp,sr = tst.SpearmansCorrelation()
#print tissue, sample
#if rho>.5: print [rho, pr, sr],[pp,sp];sys.exit()
#if rho<.5: print [rho, pr, sr],[pp,sp];sys.exit()
sample_exp_db=[]; tissue_exp_db=[]
#print 'Correlation analysis finished'
def pearson(array1,array2):
item = 0; sum_a = 0; sum_b = 0; sum_c = 0
while item < len(array1):
a = (array1[item] - avg(array1))*(array2[item] - avg(array2))
b = math.pow((array1[item] - avg(array1)),2)
c = math.pow((array2[item] - avg(array2)),2)
sum_a = sum_a + a
sum_b = sum_b + b
sum_c = sum_c + c
item = item + 1
r = sum_a/math.sqrt(sum_b*sum_c)
return r
def avg(array):
try: return sum(array)/len(array)
except Exception: return 0
def adjustPValues():
""" Can be applied to calculate an FDR p-value on the p-value reported by scipy.
Currently this method is not employed since the p-values are not sufficiently
stringent or appropriate for this type of analysis """
import statistics
all_sample_data={}
for tissue in tissue_comparison_scores:
for (r,p,sample) in tissue_comparison_scores[tissue]:
all_sample_data[sample] = db = {} ### populate this dictionary and create sub-dictionaries
break
for tissue in tissue_comparison_scores:
for (r,p,sample) in tissue_comparison_scores[tissue]:
gs = statistics.GroupStats('','',p)
all_sample_data[sample][tissue] = gs
for sample in all_sample_data:
statistics.adjustPermuteStats(all_sample_data[sample])
for tissue in tissue_comparison_scores:
scores = []
for (r,p,sample) in tissue_comparison_scores[tissue]:
p = all_sample_data[sample][tissue].AdjP()
scores.append([r,p,sample])
tissue_comparison_scores[tissue] = scores
def stdev(array):
sum_dev = 0
x_bar = scipy.average(array)
n = float(len(array))
for x in array:
x = float(x)
sq_deviation = math.pow((x-x_bar),2)
sum_dev += sq_deviation
try:
s_sqr = (1.0/(n-1.0))*sum_dev #s squared is the variance
s = math.sqrt(s_sqr)
except ZeroDivisionError:
s = 'null'
return s
def replacePearsonPvalueWithZscore():
all_sample_data={}
for tissue in tissue_comparison_scores:
for (r,p,sample) in tissue_comparison_scores[tissue]:
all_sample_data[sample] = [] ### populate this dictionary and create sub-dictionaries
break
for tissue in tissue_comparison_scores:
for (r,p,sample) in tissue_comparison_scores[tissue]:
all_sample_data[sample].append(r)
sample_stats={}
all_dataset_rho_values=[]
### Get average and standard deviation for all sample rho's
for sample in all_sample_data:
all_dataset_rho_values+=all_sample_data[sample]
avg=scipy.average(all_sample_data[sample])
st_dev=stdev(all_sample_data[sample])
sample_stats[sample]=avg,st_dev
global_rho_avg = scipy.average(all_dataset_rho_values)
global_rho_stdev = stdev(all_dataset_rho_values)
### Replace the p-value for each rho
for tissue in tissue_comparison_scores:
scores = []
for (r,p,sample) in tissue_comparison_scores[tissue]:
#u,s=sample_stats[sample]
#z = (r-u)/s
z = (r-global_rho_avg)/global_rho_stdev ### Instead of doing this for the sample background, do it relative to all analyzed samples
scores.append([r,z,sample])
tissue_comparison_scores[tissue] = scores
def exportCorrelationResults():
corr_output_file = string.replace(exp_output_file,'DATASET','LineageCorrelations')
corr_output_file = string.replace(corr_output_file,'.txt','-'+coding_type+'.txt')
if analysis_type == 'AltExon':
corr_output_file = string.replace(corr_output_file,coding_type,'AltExon')
filename = string.split(corr_output_file,delim)[-1][:-4]
#score_data = exportFile(corr_output_file)
if use_scipy:
zscore_output_dir = string.replace(corr_output_file,'.txt','-zscores.txt')
probability_data = exportFile(zscore_output_dir)
#adjustPValues()
replacePearsonPvalueWithZscore()
### Make title row
headers=['Sample_name']
for tissue in tissue_comparison_scores:
for (r,z,sample) in tissue_comparison_scores[tissue]: headers.append(sample)
break
title_row = string.join(headers,'\t')+'\n'
#score_data.write(title_row)
if use_scipy:
probability_data.write(title_row)
### Export correlation data
tissue_scores = {}; tissue_probabilities={}; tissue_score_list = [] ### store and rank tissues according to max(score)
for tissue in tissue_comparison_scores:
scores=[]
probabilities=[]
for (r,z,sample) in tissue_comparison_scores[tissue]:
scores.append(r)
probabilities.append(z)
tissue_score_list.append((max(scores),tissue))
tissue_scores[tissue] = probabilities ### These are actually z-scores
#tissue_scores[tissue] = string.join(map(str,[tissue]+scores),'\t')+'\n' ### export line
if use_scipy:
tissue_probabilities[tissue] = string.join(map(str,[tissue]+probabilities),'\t')+'\n'
tissue_score_list.sort()
tissue_score_list.reverse()
#for (score,tissue) in tissue_score_list:
#score_data.write(tissue_scores[tissue])
#if use_scipy: probability_data.write(tissue_probabilities[tissue])
#score_data.close()
if use_scipy:
probability_data.close()
#print filename,'exported...'
tissue_scores['headers'] = headers
return zscore_output_dir, tissue_scores
def visualizeLineageZscores(zscore_output_dir,grouped_lineage_zscore_dir,graphic_links):
import clustering
### Perform hierarchical clustering on the LineageProfiler Zscores
graphic_links = clustering.runHCOnly(zscore_output_dir,graphic_links)
return graphic_links
if __name__ == '__main__':
################ Default Variables ################
species = 'Hs'
platform = "exon"
vendor = 'Affymetrix'
compendium_platform = "exon"
codingtype = 'protein_coding'
platform = vendor, platform
exp_output = None
geneModels = False
modelSize = None
""" This script iterates the LineageProfiler algorithm (correlation based classification method) to identify sample types relative
two one of two references given one or more gene models. The program '
"""
#python LineageProfilerIterate.py --i "/Users/nsalomonis/Desktop/dataAnalysis/Sarwal/qPCR/deltaCT/LabMeeting/ExpressionInput/exp.ABI_PediatricSNS.txt" --r "/Users/nsalomonis/Desktop/dataAnalysis/Sarwal/qPCR/deltaCT/LabMeeting/ExpressionOutput/MarkerFinder/MarkerFinder-ABI_PediatricSNS.txt" --m "/Users/nsalomonis/Desktop/dataAnalysis/Sarwal/qPCR/deltaCT/LabMeeting/ExpressionInput/7GeneModels.txt"
#python LineageProfilerIterate.py --i "/Users/nsalomonis/Desktop/dataAnalysis/Sarwal/qPCR/deltaCT/LabMeeting/ExpressionInput/exp.ABI_PediatricSNS.txt" --r "/Users/nsalomonis/Desktop/dataAnalysis/Sarwal/qPCR/deltaCT/LabMeeting/ExpressionOutput/MarkerFinder/MarkerFinder-ABI_PediatricSNS.txt" --s 4
################ Comand-line arguments ################
if len(sys.argv[1:])<=1: ### Indicates that there are insufficient number of command-line arguments
print "Warning! Please designate a tab-delimited input expression file in the command-line"
print 'Example: python LineageProfilerIterate.py --i "/Users/me/qPCR.txt" --r "/Users/me/reference.txt" --m "/Users/me/models.txt"'
else:
options, remainder = getopt.getopt(sys.argv[1:],'', ['i=','species=','o=','platform=','codingtype=',
'compendium_platform=','r=','m=','v=','s='])
for opt, arg in options:
if opt == '--i': exp_input=arg
elif opt == '--o': exp_output=arg
elif opt == '--platform': platform=arg
elif opt == '--codingtype': codingtype=arg
elif opt == '--compendium_platform': compendium_platform=arg
elif opt == '--r': customMarkers=arg
elif opt == '--m': geneModels=arg
elif opt == '--v': vendor=arg
elif opt == '--s':
try: modelSize = int(arg)
except Exception:
modelSize = arg
if modelSize != 'optimize':
print 'Please specify a modelSize (e.g., 7-gene model search) as a single integer (e.g., 7)'
sys.exit()
else:
print "Warning! Command-line argument: %s not recognized. Exiting..." % opt; sys.exit()
if exp_output == None: exp_output = exp_input
runLineageProfiler(species,platform,exp_input,exp_output,codingtype,compendium_platform,modelSize=modelSize,customMarkers=customMarkers,geneModels=geneModels)