def test_anova_summary():
an = ANOVA(ic50_test)
# by default regression includes + msi + feature
drug_id = 999
df = an.anova_one_drug_one_feature(drug_id, "ASH1L_mut")
x = an.anova_pvalues
x = [x["tissue"], x["msi"], x["feature"]]
y = [3.210453608523738e-06, 0.14579091345305398, 0.5430736275249095]
assert_list_almost_equal(x, y, deltas=1e-10)
an.settings.analysis_type = "COREAD" # something different from PANCAN
df = an.anova_one_drug_one_feature(drug_id, "ASH1L_mut")
x = an.anova_pvalues
x = [x["msi"], x["feature"]]
y = [0.262294448831941, 0.30599483315087317]
assert_list_almost_equal(x, y, deltas=1e-10)
# now remove also the MSI factor, in which case the tissue must also be
# removed !
an.settings.include_MSI_factor = False
an.settings.analysis_type = "COREAD"
df = an.anova_one_drug_one_feature(drug_id, "ASH1L_mut")
x = [an.anova_pvalues["feature"]]
y = [0.21266050833611852]
assert_list_almost_equal(x, y, deltas=1e-10)
assert (df.N_FEATURE_neg == 365).all()
def test_anova_summary():
an = ANOVA(ic50_test)
# by default regression includes + msi + feature
drug_id = 'Drug_999_IC50'
df = an.anova_one_drug_one_feature(drug_id, 'ASH1L_mut')
x = an.anova_pvalues
y = [3.210453608523738e-06, 0.14579091345305398, 0.5430736275249095, None]
assert_list_almost_equal(x, y)
an.settings.analysis_type = 'COREAD' # something different from PANCAN
df = an.anova_one_drug_one_feature(drug_id, 'ASH1L_mut')
x = an.anova_pvalues
y = [0.262294448831941, 0.30599483315087317, None]
assert_list_almost_equal(x, y)
# now remove also the MSI factor
an.settings.include_MSI_factor = False
df = an.anova_one_drug_one_feature(drug_id, 'ASH1L_mut')
x = an.anova_pvalues
y = [0.21266050833611852, None]
assert_list_almost_equal(x, y)
assert (df.N_FEATURE_neg == 365).all()
def _analyse_all(self, multicore=None):
for gf_filename in sorted(self.gf_filenames):
tcga = gf_filename.split("_")[1].split('.')[0]
print(purple('======================== Analysing %s data' % tcga))
self.mkdir(self.main_directory + os.sep + tcga)
# Computes the ANOVA
try:
self.ic50 = IC50(self.ic50_filename)
except:
print("Clustering IC50 (v18 released data ?)")
self.ic50 = IC50Cluster(self.ic50_filename, verbose=False)
an = ANOVA(self.ic50, gf_filename, self.drug_decode,
verbose=False)
if self.test is True:
an.features.df = an.features.df[an.features.df.columns[0:15]]
self.an = an
an.settings = ANOVASettings(**self.settings)
an.settings.analysis_type = tcga
an.init() # This reset the directory
results = an.anova_all(multicore=multicore)
an.settings.directory = self.main_directory + os.sep + tcga
# Store the results
self.results[tcga] = results
print('Analysing %s data and creating images' % tcga)
self.report = ANOVAReport(an)
self.report.settings.savefig = True
self.report.create_html_pages(onweb=False)
def test_anova_one_drug_one_feature():
an = ANOVA(ic50_test)
# test 1 drug
drug_id = 'Drug_999_IC50'
df = an.anova_one_drug_one_feature(
drug_id=drug_id,
feature_name='ABCB1_mut', show=True)
control = {'DRUG_ID': {1: drug_id},
'DRUG_NAME': {1: drug_id},
'DRUG_TARGET': {1: drug_id},
'FEATURE': {1: 'ABCB1_mut'},
'ANOVA_FEATURE_pval': {1: 0.86842684367357359},
'FEATURE_IC50_T_pval': {1: 0.48586107208790896},
'FEATURE_IC50_effect_size': {1: 0.31407773405409201},
'FEATURE_delta_MEAN_IC50': {1: -0.36105662590553411},
'FEATUREneg_Glass_delta': {1: 0.31296252976074801},
'FEATUREneg_IC50_sd': {1: 1.1536736560173895},
'FEATUREneg_logIC50_MEAN': {1: 2.8007757068403043},
'FEATUREpos_Glass_delta': {1: 0.53754504818376181},
'FEATUREpos_IC50_sd': {1: 0.67167696386648634},
'FEATUREpos_logIC50_MEAN': {1: 2.4397190809347702},
'ANOVA_MSI_pval': {1: 0.14598946672374763},
'N_FEATURE_neg': {1: 370},
'N_FEATURE_pos': {1: 5},
'ANOVA_TISSUE_pval': {1: 3.2808255732569986e-06}}
control = pd.DataFrame(control)
assert_list_almost_equal(df,control)
def test_multicore():
an = ANOVA(ic50_test)
results = an.anova_all()
an2 = ANOVA(ic50_test)
results2 = an2.anova_all(multicore=2)
all(results.df.fillna(0) == results2.df.fillna(0))
def test_anova_all():
an = ANOVA(ic50_test)
# slow, let us cut the features to keep only tenish
# this is a trick but would be nice to have this in the API
features = an.features.df
features = features[features.columns[0:12]]
an = ANOVA(ic50_test, features)
results = an.anova_all()
assert_almost_equal( results.df['ANOVA_FEATURE_FDR'].sum(),
10312.23061065521, 6)
an2 = ANOVA(ic50_test, features, set_media_factor=True)
results = an2.anova_all()
assert_almost_equal( results.df['ANOVA_FEATURE_FDR'].sum(),
10238.529313503008, 6)
def test_anova_all():
an = ANOVA(ic50_test)
# slow, let us cut the features to keep only ten-ish values
# this is a trick but would be nice to have this in the API
features = an.features.df
features = features[features.columns[0:12]]
an = ANOVA(ic50_test, features)
results = an.anova_all()
results # test __repr__
results.volcano()
results.barplot_effect_size()
assert_almost_equal(results.df["ANOVA_FEATURE_FDR"].sum(), 10312.23061065521, delta=1e-6)
an2 = ANOVA(ic50_test, features, set_media_factor=True)
results = an2.anova_all()
assert_almost_equal(results.df["ANOVA_FEATURE_FDR"].sum(), 10238.529313503008, delta=1e-6)
def multicore_anova(ic50, genomic_features, drug_decode=None, maxcpu=2,
sampling=0):
"""Using 4 cores, the entire analysis took 15 minutes using
4 CPUs (16 Oct 2015).
:param ic50: a filename or :class:`IC50` instance.
:return: the anova instance itself (not the results); see example below.
::
from gdsctool qq qs.anova import multicore
master = multicore(dataset, maxcpu=2)
results = master.anova_all()
from gdsctools import ANOVAReport()
report = ANOVAReport(master, results)
report.create_html_pages(0
.. warning:: experimental. Seems to work but sometimes hangs forever.
"""
print("experimental code to run the analysis with several cores")
print("May takes lots or resources and slow down your system")
t1 = time.time()
master = ANOVA(ic50, genomic_features=genomic_features,
drug_decode=drug_decode, low_memory=True)
master.sampling = sampling
drugs = master.ic50.drugIds
t = MultiProcessing(maxcpu=maxcpu)
# add all jobs (one per drug)
for i, drug in enumerate(drugs):
t.add_job(analyse_one_drug, master, drug)
t.run()
# populate the ANOVA instance with the results
for this in t.results:
drug = this[0]
result = this[1]
master.individual_anova[drug] = result
print("\nTook " + str(time.time() - t1) + "seconds.")
return master
def test_html():
# same as above, could factorise
an = ANOVA(ic50_test)
features = an.features.df
features = features[features.columns[0:30]]
an = ANOVA(ic50_test, features)
#an.settings.include_media_factor = False
results = an.anova_all()
assert len(results.df) == 302
r = ANOVAReport(gdsc=an, results=results)
assert len(r.get_significant_set()) == 3
# long but should cover everthinh.
try:
r.create_html_pages()
except Exception as err:
raise err
finally:
import shutil
shutil.rmtree('html_gdsc_anova')
def test_anova_one_drug_one_feature():
an = ANOVA(ic50_test)
# test 1 drug
drug_id = 999
df = an.anova_one_drug_one_feature(drug_id=drug_id, feature_name="ABCB1_mut", show=True)
df["DRUG_ID"] = drug_id
df["DRUG_NAME"] = drug_id
df["DRUG_TARGET"] = drug_id
df["ANOVA_MEDIA_pval"] = -1
control = {
"DRUG_ID": {1: drug_id},
"DRUG_NAME": {1: drug_id},
"DRUG_TARGET": {1: drug_id},
"FEATURE": {1: "ABCB1_mut"},
"ANOVA_FEATURE_pval": {1: 0.86842684367357359},
"FEATURE_IC50_T_pval": {1: 0.48586107208790896},
"FEATURE_IC50_effect_size": {1: 0.31407773405409201},
"FEATURE_delta_MEAN_IC50": {1: -0.36105662590553411},
# 'FEATUREneg_Glass_delta': {1: 0.31296252976074801},
# 'FEATUREneg_IC50_sd': {1: 1.1536736560173895},
"FEATUREneg_logIC50_MEAN": {1: 2.8007757068403043},
# 'FEATUREpos_Glass_delta': {1: 0.53754504818376181},
# 'FEATUREpos_IC50_sd': {1: 0.67167696386648634},
"FEATUREpos_logIC50_MEAN": {1: 2.4397190809347702},
"ANOVA_MSI_pval": {1: 0.14598946672374763},
"N_FEATURE_neg": {1: 370},
"N_FEATURE_pos": {1: 5},
"ANOVA_MEDIA_pval": {1: -1},
"ANOVA_TISSUE_pval": {1: 3.2808255732569986e-06},
}
control = pd.DataFrame(control)
for this in control.columns:
if this in ["FEATURE_delta_MEAN_IC50", "FEATUREneg_logIC50_MEAN", "FEATUREpos_logIC50_MEAN"]:
pass
else:
assert_almost_equal(df[this].values[0], control[this].values[0])
def _analyse_all(self):
for gf_filename in sorted(self.gf_filenames):
tcga = gf_filename.split("_")[1].split('.')[0]
print('================================ Analysing %s data' % tcga)
self.mkdir('ALL' + os.sep + tcga)
# Computes the ANOVA
an = ANOVA(self.ic50_filename, gf_filename, self.drug_decode)
self.an = an
an.settings = ANOVASettings(**self.settings)
an.init() # reset the analysis_type automatically
results = an.anova_all()
# Store the results
self.results[tcga] = results
print('Analysing %s data and creating images' % tcga)
self.report = ANOVAReport(an, self.results[tcga])
self.report.settings.savefig = True
self.report.settings.directory = 'ALL/' + tcga
self.report.settings.analysis_type = tcga
self.report.create_html_pages()
def test_odof_with_without_media():
gdsc = ANOVA(ic50_test)
_res = gdsc.anova_one_drug_one_feature('Drug_1047_IC50', 'TP53_mut')
dd1 = gdsc._get_anova_summary(gdsc.data_lm, output='dict')
assert_list_almost_equal([dd1['feature'], dd1['msi'], dd1['tissue']],
[1.5750735472022118e-58, 0.025902887791637515,
5.541879283763767e-44])
gdsc = ANOVA(ic50_test, set_media_factor=True)
_res = gdsc.anova_one_drug_one_feature('Drug_1047_IC50', 'TP53_mut')
dd2 = gdsc._get_anova_summary(gdsc.data_lm, output='dict')
assert_list_almost_equal([dd2['feature'], dd2['media'], dd2['msi'],
dd2['tissue']], [ 2.9236500715529455e-58, 0.7762487502315283,
0.023777744527686766, 1.5729157319290974e-44])
def test_odof_with_without_media():
gdsc = ANOVA(ic50_test)
_res = gdsc.anova_one_drug_one_feature(1047, "TP53_mut")
odof = gdsc._get_one_drug_one_feature_data(1047, "TP53_mut")
dd1 = gdsc._get_anova_summary(gdsc.data_lm, output="dict", odof=odof)
assert_almost_equal(dd1["feature"], 1.5750735472022118e-58, delta=1e-64)
assert_almost_equal(dd1["msi"], 0.025902887791637515, delta=1e-10)
# Change in Nov 2016
# assert_almost_equal(dd1['tissue'], 5.541879283763767e-44, delta=1e-50)
assert_almost_equal(dd1["tissue"], 1.0258702741509e-44, delta=1e-50)
gdsc = ANOVA(ic50_test, set_media_factor=True)
_res = gdsc.anova_one_drug_one_feature(1047, "TP53_mut")
odof = gdsc._get_one_drug_one_feature_data(1047, "TP53_mut")
dd2 = gdsc._get_anova_summary(gdsc.data_lm, output="dict", odof=odof)
assert_almost_equal(dd2["feature"], 2.9236500715529455e-58, delta=1e-64)
assert_almost_equal(dd2["media"], 0.7762487502315283, delta=1e-10)
assert_almost_equal(dd2["msi"], 0.023777744527686766, delta=1e-10)
assert_almost_equal(dd2["tissue"], 1.5729157319290974e-44, delta=1e-50)
def test_compare_formula_vs_gdsc():
# TISSUE + MSI + feature
an = ANOVA(ic50_test)
drug_id = 999
df1 = an.anova_one_drug_one_feature(drug_id=drug_id, feature_name="ABCB1_mut")
an.settings.regression_formula = "Y ~ C(tissue) + C(msi) + feature"
df2 = an.anova_one_drug_one_feature(drug_id=drug_id, feature_name="ABCB1_mut")
assert all(df1.fillna(1).values[0] == df2.fillna(1).values[0])
# MSI + feature only
an = ANOVA(ic50_test)
an.settings.analysis_type = "breast"
df1 = an.anova_one_drug_one_feature(drug_id=drug_id, feature_name="ABCB1_mut")
an.settings.regression_formula = "Y ~ C(msi) + feature"
df2 = an.anova_one_drug_one_feature(an.drugIds[2], an.feature_names[0])
assert all(df1.fillna(1) == df2.fillna(1))
def test_set_cancer_type():
an = ANOVA(gdsctools_data("IC50_v17.csv.gz"))
an.set_cancer_type("breast")
assert_list_almost_equal([an.ic50.df.sum().sum()], [27721.255627472943])
def test_anova_one_drug_one_feature():
an = ANOVA(ic50_test)
# test 1 drug
drug_id = 'Drug_999_IC50'
df = an.anova_one_drug_one_feature(drug_id=drug_id,
feature_name='ABCB1_mut',
show=True)
control = {
'DRUG_ID': {
1: drug_id
},
'DRUG_NAME': {
1: drug_id
},
'DRUG_TARGET': {
1: drug_id
},
'FEATURE': {
1: 'ABCB1_mut'
},
'ANOVA_FEATURE_pval': {
1: 0.86842684367357359
},
'FEATURE_IC50_T_pval': {
1: 0.48586107208790896
},
'FEATURE_IC50_effect_size': {
1: 0.31407773405409201
},
'FEATURE_delta_MEAN_IC50': {
1: -0.36105662590553411
},
'FEATUREneg_Glass_delta': {
1: 0.31296252976074801
},
'FEATUREneg_IC50_sd': {
1: 1.1536736560173895
},
'FEATUREneg_logIC50_MEAN': {
1: 2.8007757068403043
},
'FEATUREpos_Glass_delta': {
1: 0.53754504818376181
},
'FEATUREpos_IC50_sd': {
1: 0.67167696386648634
},
'FEATUREpos_logIC50_MEAN': {
1: 2.4397190809347702
},
'ANOVA_MSI_pval': {
1: 0.14598946672374763
},
'N_FEATURE_neg': {
1: 370
},
'N_FEATURE_pos': {
1: 5
},
'ANOVA_TISSUE_pval': {
1: 3.2808255732569986e-06
}
}
control = pd.DataFrame(control)
assert_list_almost_equal(df, control)
def test_anova_one_drug():
# test entire drug across all fearures
an = ANOVA(ic50_test)
df = an.anova_one_drug('Drug_999_IC50')
def test_anova_one_drug():
# test entire drug across all fearures
an = ANOVA(ic50_test)
df = an.anova_one_drug('Drug_999_IC50')
def create_data_packages_for_companies(self, companies=None):
"""Creates a data package for each company found in the DrugDecode file
"""
##########################################################
#!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!#
# #
# DRUG_DECODE and IC50 inputs must be filtered to keep #
# only WEBRELEASE=Y and owner #
# #
#!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!#
##########################################################
# companies must be just one name (one string) or a list of strings
# By default, takes all companies found in DrugDecode
if isinstance(companies, str):
companies = [companies]
if companies is None:
companies = self.companies
if len(companies) == 0:
raise ValueError("Could not find any companies in the DrugDecode file")
# The main directory
self.mkdir(self.company_directory)
# Loop over all companies, retrieving information built
# in analyse() method, selecting for each TCGA all information
# for that company only (and public drugs)
Ncomp = len(companies)
for ii, company in enumerate(companies):
print(purple("\n=========== Analysing company %s out of %s (%s)" %
(ii+1, Ncomp, company)))
self.mkdir(self.company_directory + os.sep + company)
# Handle each TCGA case separately
for gf_filename in sorted(self.gf_filenames):
tcga = gf_filename.split("_")[1].split('.')[0]
print(brown(" ------- building TCGA %s sub directory" % tcga))
# Read the results previously computed either
try:
results_df = self.results[tcga].df.copy()
except:
results_path = "%s/%s/OUTPUT/results.csv" % (self.main_directory, tcga)
results_df = ANOVAResults(results_path)
# MAke sure the results are formatted correctly
results = ANOVAResults(results_df)
# Get the DrugDecode information for that company only
drug_decode_company = self.drug_decode.df.query(
"WEBRELEASE=='Y' or OWNED_BY=='%s'" % company)
# Transform into a proper DrugDecode class for safety
drug_decode_company = DrugDecode(drug_decode_company)
# Filter the results to keep only public drugs and that
# company. Make sure this is integers
results.df["DRUG_ID"] = results.df["DRUG_ID"].astype(int)
mask = [True if x in drug_decode_company.df.index else False
for x in results.df.DRUG_ID]
results.df = results.df.ix[mask]
# We read the IC50 again
try:
self.ic50 = IC50(self.ic50_filename)
except:
self.ic50 = IC50Cluster(self.ic50_filename, verbose=False)
# And create an ANOVA instance. This is not to do the analyse
# again but to hold various information
an = ANOVA(self.ic50, gf_filename, drug_decode_company,
verbose=False)
def drug_to_keep(drug):
to_keep = drug in drug_decode_company.df.index
return to_keep
an.ic50.df = an.ic50.df.select(drug_to_keep, axis=1)
an.settings = ANOVASettings(**self.settings)
an.init()
an.settings.directory = self.company_directory + os.sep + company + os.sep + tcga
an.settings.analysis_type = tcga
# Now we create the report
self.report = ANOVAReport(an, results,
drug_decode=drug_decode_company,
verbose=self.verbose)
self.report.company = company
self.report.settings.analysis_type = tcga
self.report.create_html_main(False)
self.report.create_html_manova(False)
self.report.create_html_features()
self.report.create_html_drugs()
self.report.create_html_associations()
def create_data_packages_for_companies(self, companies=None):
##########################################################
#!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!#
# #
# DRUG_DECODE and IC50 inputs must be filtered to keep #
# only WEBRELEASE=Y and owner #
# #
#!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!#
##########################################################
if isinstance(companies, str):
companies = [companies]
if companies is None:
companies = self.companies
Ncomp = len(companies)
for ii, company in enumerate(companies):
print("\n\n========= Analysing company %s out of %s (%s)" %
(ii + 1, Ncomp, company))
self.mkdir(company)
for gf_filename in sorted(self.gf_filenames):
tcga = gf_filename.split("_")[1].split('.')[0]
print("---------------- for TCGA %s" % tcga)
# Read the results previously computed
try:
results_df = self.results[tcga].df.copy()
except:
results_path = "ALL/%s/OUTPUT/results.csv" % tcga
print("Downloading results from %s" % results_path)
results_df = ANOVAResults(results_path)
results = ANOVAResults(results_df)
# Get a DrugDecode for that company
drug_decode_company = self.drug_decode.df.query(
"WEBRELEASE=='Y' or OWNED_BY=='%s'" % company)
# Transform into a proper DrugDecode class for safety
drug_decode_company = DrugDecode(drug_decode_company)
# filter results using the new drug decode
drug_ids_in_results = get_drug_id(results.df.DRUG_ID)
mask = [
True if x in drug_decode_company.df.index else False
for x in drug_ids_in_results
]
results.df = results.df.ix[mask]
# Just to create an instance with the subset of drug_decode
# and correct settings. This is also used to store
# the entire input data set. So, we must remove all drugs
# not relevant for the analysis of this company
an = ANOVA(self.ic50_filename, gf_filename,
drug_decode_company)
def drug_to_keep(drug):
to_keep = get_drug_id(drug) in drug_decode_company.df.index
return to_keep
an.ic50.df = an.ic50.df.select(drug_to_keep, axis=1)
an.settings = ANOVASettings(**self.settings)
an.init()
an.settings.directory = company + os.sep + tcga
an.settings.analysis_type = tcga
self.report = ANOVAReport(an, results)
self.report.settings.analysis_type = tcga
self.report.create_html_main(False)
self.report.create_html_manova(False)
if self.debug is False:
self.report.create_html_features()
self.report.create_html_associations()
# For now, we just copy all DRUG images from
# the analysis made in ALL
from easydev import shellcmd, Progress
print("\nCopying drug files")
drug_ids = results.df.DRUG_ID.unique()
pb = Progress(len(drug_ids))
for i, drug_id in enumerate(drug_ids):
# copy the HTML
filename = "%s.html" % drug_id
source = "ALL%s%s%s" % (os.sep, tcga, os.sep)
dest = "%s%s%s%s" % (company, os.sep, tcga, os.sep)
cmd = "cp %s%s %s" % (source, filename, dest)
shellcmd(cmd, verbose=False)
#copy the images
filename = "volcano_%s.*" % drug_id
source = "ALL%s%s%simages%s" % (os.sep, tcga, os.sep,
os.sep)
dest = "%s%s%s%simages%s" % (company, os.sep, tcga,
os.sep, os.sep)
cmd = "cp %s%s %s" % (source, filename, dest)
shellcmd(cmd, verbose=False)
pb.animate(i + 1)
def create_data_packages_for_companies(self, companies=None):
##########################################################
#!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!#
# #
# DRUG_DECODE and IC50 inputs must be filtered to keep #
# only WEBRELEASE=Y and owner #
# #
#!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!#
##########################################################
if isinstance(companies, str):
companies = [companies]
if companies is None:
companies = self.companies
Ncomp = len(companies)
for ii, company in enumerate(companies):
print("\n\n========= Analysing company %s out of %s (%s)" %
(ii+1, Ncomp, company))
self.mkdir(company)
for gf_filename in sorted(self.gf_filenames):
tcga = gf_filename.split("_")[1].split('.')[0]
print("---------------- for TCGA %s" % tcga)
# Read the results previously computed
try:
results_df = self.results[tcga].df.copy()
except:
results_path = "ALL/%s/OUTPUT/results.csv" % tcga
print("Downloading results from %s" % results_path)
results_df = ANOVAResults(results_path)
results = ANOVAResults(results_df)
# Get a DrugDecode for that company
drug_decode_company = self.drug_decode.df.query(
"WEBRELEASE=='Y' or OWNED_BY=='%s'" % company)
# Transform into a proper DrugDecode class for safety
drug_decode_company = DrugDecode(drug_decode_company)
# filter results using the new drug decode
drug_ids_in_results = get_drug_id(results.df.DRUG_ID)
mask = [True if x in drug_decode_company.df.index else False
for x in drug_ids_in_results]
results.df = results.df.ix[mask]
# Just to create an instance with the subset of drug_decode
# and correct settings. This is also used to store
# the entire input data set. So, we must remove all drugs
# not relevant for the analysis of this company
an = ANOVA(self.ic50_filename, gf_filename, drug_decode_company)
def drug_to_keep(drug):
to_keep = get_drug_id(drug) in drug_decode_company.df.index
return to_keep
an.ic50.df = an.ic50.df.select(drug_to_keep, axis=1)
an.settings = ANOVASettings(**self.settings)
an.init()
an.settings.directory = company + os.sep + tcga
an.settings.analysis_type = tcga
self.report = ANOVAReport(an, results)
self.report.settings.analysis_type = tcga
self.report.create_html_main(False)
self.report.create_html_manova(False)
if self.debug is False:
self.report.create_html_features()
self.report.create_html_associations()
# For now, we just copy all DRUG images from
# the analysis made in ALL
from easydev import shellcmd, Progress
print("\nCopying drug files")
drug_ids = results.df.DRUG_ID.unique()
pb = Progress(len(drug_ids))
for i, drug_id in enumerate(drug_ids):
# copy the HTML
filename = "%s.html" % drug_id
source = "ALL%s%s%s" % (os.sep, tcga, os.sep)
dest = "%s%s%s%s" % (company, os.sep, tcga, os.sep )
cmd = "cp %s%s %s" % (source, filename, dest )
shellcmd(cmd, verbose=False)
#copy the images
filename = "volcano_%s.*" % drug_id
source = "ALL%s%s%simages%s" % (os.sep, tcga,
os.sep, os.sep)
dest = "%s%s%s%simages%s" % (company, os.sep,
tcga, os.sep , os.sep)
cmd = "cp %s%s %s" % (source, filename, dest )
shellcmd(cmd, verbose=False)
pb.animate(i+1)
def create_data_packages_for_companies(self, companies=None):
"""Creates a data package for each company found in the DrugDecode file
"""
##########################################################
#!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!#
# #
# DRUG_DECODE and IC50 inputs must be filtered to keep #
# only WEBRELEASE=Y and owner #
# #
#!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!#
##########################################################
# companies must be just one name (one string) or a list of strings
# By default, takes all companies found in DrugDecode
if isinstance(companies, str):
companies = [companies]
if companies is None:
companies = self.companies
if len(companies) == 0:
raise ValueError(
"Could not find any companies in the DrugDecode file")
# The main directory
self.mkdir(self.company_directory)
# Loop over all companies, retrieving information built
# in analyse() method, selecting for each TCGA all information
# for that company only (and public drugs)
Ncomp = len(companies)
for ii, company in enumerate(companies):
print(
purple("\n=========== Analysing company %s out of %s (%s)" %
(ii + 1, Ncomp, company)))
self.mkdir(self.company_directory + os.sep + company)
# Handle each TCGA case separately
for gf_filename in sorted(self.gf_filenames):
tcga = gf_filename.split("_")[1].split('.')[0]
print(brown(" ------- building TCGA %s sub directory" % tcga))
# Read the results previously computed either
try:
results_df = self.results[tcga].df.copy()
except:
results_path = "%s/%s/OUTPUT/results.csv" % (
self.main_directory, tcga)
results_df = ANOVAResults(results_path)
# MAke sure the results are formatted correctly
results = ANOVAResults(results_df)
# Get the DrugDecode information for that company only
drug_decode_company = self.drug_decode.df.query(
"WEBRELEASE=='Y' or OWNED_BY=='%s'" % company)
# Transform into a proper DrugDecode class for safety
drug_decode_company = DrugDecode(drug_decode_company)
# Filter the results to keep only public drugs and that
# company. Make sure this is integers
results.df["DRUG_ID"] = results.df["DRUG_ID"].astype(int)
mask = [
True if x in drug_decode_company.df.index else False
for x in results.df.DRUG_ID
]
results.df = results.df.ix[mask]
# We read the IC50 again
try:
self.ic50 = IC50(self.ic50_filename)
except:
self.ic50 = IC50Cluster(self.ic50_filename, verbose=False)
# And create an ANOVA instance. This is not to do the analyse
# again but to hold various information
an = ANOVA(self.ic50,
gf_filename,
drug_decode_company,
verbose=False)
def drug_to_keep(drug):
to_keep = drug in drug_decode_company.df.index
return to_keep
an.ic50.df = an.ic50.df.select(drug_to_keep, axis=1)
an.settings = ANOVASettings(**self.settings)
an.init()
an.settings.directory = self.company_directory + os.sep + company + os.sep + tcga
an.settings.analysis_type = tcga
# Now we create the report
self.report = ANOVAReport(an,
results,
drug_decode=drug_decode_company,
verbose=self.verbose)
self.report.company = company
self.report.settings.analysis_type = tcga
self.report.create_html_main(False)
self.report.create_html_manova(False)
self.report.create_html_features()
self.report.create_html_drugs()
self.report.create_html_associations()
def test_set_cancer_type():
an = ANOVA(gdsctools_data("IC50_v17.csv.gz"))
an.set_cancer_type("breast")
assert_list_almost_equal([an.ic50.df.sum().sum()], [27721.255627472943])
