netZooPy
netZooPy copied to clipboard
netZoo
LionessPuma issue: compute_puma() missing 1 required positional argument: 'sorted_index'
from netZooPy.panda.panda import Panda
from netZooPy.puma.puma import Puma
from netZooPy.lioness.lioness import Lioness
from netZooPy.lioness.lioness_for_puma import LionessPuma
import pandas as pd
import os
from io import StringIO
import sys
class Capturing(list):
def __enter__(self):
self._stdout = sys.stdout
sys.stdout = self._stringio = StringIO()
return self
def __exit__(self, *args):
self.extend(self._stringio.getvalue().splitlines())
del self._stringio # free up some memory
sys.stdout = self._stdout
def read_expression_file(filepath = '',header = 'infer',server = 'PANDA',sample_mode = ''):
header_float= False
df = pd.read_csv(filepath,delimiter = ',',index_col = 0,header = 'infer')
if server == 'LIONESS':
if sample_mode == 'sample_name':
return df
elif sample_mode == "sample_num":
try:
header_int = [int(col) for col in df.columns]
header_float = [float(col) for col in df.columns]
sample_list = list(range(len(df.columns)))
if sample_list == header_int or sample_list == list(df.columns):
return df
df = pd.read_csv(filepath,index_col =0 ,sep = ',', header = None)
return df
except:
return df
try:
header_float = [float(col) for col in df.columns]
header_float = True
except:
header_float = False
if header_float:
df = pd.read_csv(filepath,index_col =0 ,sep = ',', header = None)
return df
def create_panda_obj(dataframes = {} , filepaths = {},alpha = 0.1,mode = 'union',precision = "double"):
#print("panda_input_filepaths: ",filepaths)
panda_obj = None
filepaths_values = list(filepaths.values())
if isinstance(filepaths_values,list):
if len(filepaths_values) == 3:
if filepaths_values[0] == "" or filepaths_values[1] == "" or filepaths_values[2] == "":
filepaths_values = []
elif len(filepaths_values) < 3:
filepaths_values = []
else:
filepaths_values = []
if len(dataframes.keys()) > 0:
panda_obj = Panda(dataframes['expression'],dataframes['motif'],dataframes['ppi'],modeProcess = mode, alpha = alpha,save_memory=False,precision=precision,remove_missing=False, keep_expression_matrix=True)
elif len(filepaths_values)>0:
panda_obj = Panda(filepaths["expression"],filepaths["motif"],filepaths["ppi"],modeProcess = mode, alpha = alpha,save_memory=False,precision=precision,remove_missing=False, keep_expression_matrix=True)
#print("panda_obj: ",panda_obj)
return panda_obj
def create_puma_obj(dataframes = {} , filepaths = {},alpha = 0.1, mode = "union",precision = "double",keep_expression_matrix= True):
#print("#######creating puma obj#####")
puma_obj = None
all_files_uploaded = True
#print(filepaths)
if len(filepaths.keys()) > 0:
puma_obj = Puma(filepaths["expression"],filepaths["motif"],None,filepaths["mir"],save_memory=False,alpha = alpha, modeProcess = mode,precision = precision,keep_expression_matrix=keep_expression_matrix,save_tmp=False)
if len(dataframes.keys()) > 0:
puma_obj = Puma(dataframes['expression'],dataframes['motif'],None,dataframes['mir'],alpha = alpha, modeProcess = mode,precision = precision,keep_expression_matrix=keep_expression_matrix,save_tmp=False)
return puma_obj
def run_lioness(dataframes = {},input_mode = 'PANDA',alpha = 0.1,mode = 'union',precision='double',start = 0 ,end = 0):
t = type('test', (object,), {})()
cwd = os.getcwd()
output = []
lioness_results_folder = os.path.join(cwd,"results","lioness")
if not os.path.isdir(lioness_results_folder):
os.makedirs(lioness_results_folder)
try:
if input_mode == "PANDA":
with Capturing() as output:
try:
panda_obj = create_panda_obj(dataframes= dataframes,alpha = float(alpha), mode = mode,precision=precision,remove_missing=False, keep_expression_matrix= True)
except:
return {'status': 'failed', 'reason': 'PANDA_ANALYSIS_ERROR'}
print("panda_obj:", panda_obj)
try:
lioness_obj = Lioness(panda_obj,save_dir = lioness_results_folder,start = start,end= end,save_fmt = "aaa",alpha = float(alpha),precision=precision)
except:
return {'status': 'failed', 'reason': 'LIONESS_ANALYSIS_ERROR'}
try:
Panda.processData(self=t,modeProcess=mode, motif_file=dataframes["motif"], expression_file=dataframes["expression"], ppi_file=dataframes["ppi"], remove_missing=False, keep_expression_matrix=False)
except:
return {'status': 'failed', 'reason': 'PANDA_ANALYSIS_ERROR'}
W = t.motif_matrix_unnormalized
tfs = t.unique_tfs
genes = t.gene_names
#print("####W######")
# #print(W)
# #print(W.shape)
W_df = pd.DataFrame(W,index = tfs,columns = genes)
W_df= W_df.stack().reset_index().rename(columns={'level_0':'tf','level_1':'gene', 0:'motif'})
lioness_obj_df = pd.DataFrame(lioness_obj.export_lioness_results)
print(lioness_obj_df)
col_names = ["tf","gene","force"]
lioness_obj_df.columns = col_names
lioness_obj_df["motif"] = W_df['motif']
cols_order = ["tf","gene","motif","force"]
lioness_obj_df = lioness_obj_df[cols_order]
adj_matrix = create_adj_matrix_from_rows_df(df = lioness_obj_df,gene_names= genes, unique_tfs= tfs,server_name = "lioness")
#print("after:",adj_matrix.shape)
#print("###adj_matrix_targeting######")
#print(adj_matrix)
lioness_adj_mat_df = pd.DataFrame(adj_matrix,index = tfs,columns = genes)
elif input_mode == "PUMA":
#puma.py require mir to be a filepath
with Capturing() as output:
puma_obj = create_puma_obj(dataframes = dataframes,alpha = alpha, mode = mode,precision = precision,keep_expression_matrix=True)
#print("puma_obj:",puma_obj)
lioness_obj = LionessPuma(puma_obj,save_dir = lioness_results_folder,start = start,end= end,save_fmt = "aaa",alpha = float(alpha),precision=precision, )
#print("lkionessPuma:",lioness_obj)
Panda.processData(self=t,modeProcess=mode, motif_file=dataframes["motif"], expression_file=dataframes["expression"], ppi_file=None, remove_missing=False, keep_expression_matrix=False)
W = t.motif_matrix_unnormalized
tfs = t.unique_tfs
genes = t.gene_names
#print("####W######")
# #print(W)
# #print(W.shape)
W_df = pd.DataFrame(W,index = tfs,columns = genes)
W_df= W_df.stack().reset_index().rename(columns={'level_0':'tf','level_1':'gene', 0:'motif'})
lioness_matrix = lioness_obj.export_lioness_results
#print("###########3LIONESS_MATRIX############")
# #print(type(lioness_matrix))
# #print(lioness_matrix)
if isinstance(lioness_matrix,np.ndarray):
lioness_obj_df = pd.DataFrame(lioness_matrix,columns = ["tf","gene","motif","force"])
adj_matrix = create_adj_matrix_from_rows_df(df = lioness_obj_df,gene_names= genes, unique_tfs= tfs,server_name = "lioness")
#print("after:",adj_matrix.shape)
#print("###adj_matrix_targeting######")
#print(adj_matrix)
lioness_adj_mat_df = pd.DataFrame(adj_matrix,index = tfs,columns = genes)
elif input_mode == "Coexpression":
with Capturing() as output:
panda_obj = create_panda_obj(dataframes = dataframes,alpha = float(alpha), mode = mode,precision=precision,remove_missing=False, keep_expression_matrix= True)
lioness_obj = Lioness(panda_obj,save_dir = lioness_results_folder,start = start,end= end,save_fmt = "aaa",alpha = float(alpha),precision=precision)
# Panda.processData(self=t,modeProcess=mode, motif_file=lioness_input_filepaths["motif"], expression_file=lioness_input_filepaths["expression"], ppi_file=None, remove_missing=False, keep_expression_matrix=False)
# W = None
# tfs = t.unique_tfs
# genes = t.gene_names
col_names = ["gene1","gene2","force"]
#setting columns = ["tf,"gene","force"] just to make the frontend work
fake_col_names = ["tf","gene","force"]
# otherwise the names should be gene1,gene2,force
lioness_obj_df = pd.DataFrame(lioness_obj.export_lioness_results)
lioness_obj_df.columns = fake_col_names
# lioness_obj_df["motif"] = -1
# cols_order = ["tf","gene","motif","force"]
cols_order = ["tf","gene","force"]
lioness_obj_df = lioness_obj_df[cols_order]
lioness_adj_mat_df = Lioness(panda_obj,save_dir = lioness_results_folder,start = start,end= end,save_fmt = "aaa",alpha = float(alpha),precision=precision,output = 'gene_targeting').export_lioness_results
except Exception as e:
print('###run_lioness_exception###')
print(e)
lioness_obj_df,lioness_adj_mat_df = None,None
return lioness_obj_df,lioness_adj_mat_df,output
def main():
folder = '/home/'
lioness_input_filepaths = {"expression": folder + 'ToyExpressionData.csv', "motif": folder + 'ToyMotifData.csv',"mir": folder + 'ToyMiRList.csv','ppi': folder + 'ToyPPIData.csv'}
mode = 'union'
precision = 'double'
start = 1
end = 1
sample_mode = 'sample_num'
input_mode = "PUMA"
sample_num = 1
alpha = 0.1
sep = ','
dataframes = {
'expression':None,
'motif': None,
'ppi': None,
'mir':None
}
dataframes['expression'] = read_expression_file(filepath = lioness_input_filepaths['expression'],header = 'infer',server = 'LIONESS',sample_mode = sample_mode)
if input_mode == "PANDA":
dataframes.update({
'motif': pd.read_csv(lioness_input_filepaths['motif'],sep = sep,header = None),
'ppi': pd.read_csv(lioness_input_filepaths['ppi'],sep = sep,header = None)
})
elif input_mode == "PUMA":
#puma.py require mir to be a filepath
dataframes.update({
'motif': pd.read_csv(lioness_input_filepaths['motif'],sep = sep,header = None),
'mir': lioness_input_filepaths['mir'],#pd.read_csv(lioness_input_filepaths['mir'],sep = sep,header = None)
})
elif input_mode == "Coexpression":
pass
else:
return {'status': 'failed', 'reason': 'INVALID_INPUT_MODE_ARGUMENT'}
lioness_obj_df,lioness_adj_mat_df,output = run_lioness(dataframes = dataframes,input_mode = input_mode,alpha = alpha,mode = mode,precision=precision,start =start ,end = end )
main()
@violafanfani @talkhanz said there is a difference between puma and lionesspuma
https://github.com/netZoo/netZooPy/blob/master/netZooPy/puma/puma.py#L319
https://github.com/netZoo/netZooPy/blob/master/netZooPy/lioness/lioness_for_puma.py#L139
