def load():
DeconRNASeq.load_r()
rpy2.robjects.r("data(rat_liver_brain)")
x = rpy2.robjects.r('all.datasets')
df = rpy2.robjects.pandas2ri.ri2py(x)
df.index = rpy2.robjects.r('rownames')(x)
sample_info = {
'd15282bm': 'liver',
'd15283bm': 'liver',
'd15284bm': 'liver',
'd15285bm': 'mix',
'd15286bm': 'mix',
'd15287bm': 'mix',
'd15288bm': 'mix',
'd15289bm': 'mix',
'd15290bm': 'mix',
'd15291bm': 'mix',
'd15292bm': 'mix',
'd15293bm': 'mix',
'd15294bm': 'brain',
'd15295bm': 'brain',
'd15296bm': 'brain',
}
return sample_info, df
def select_genes(df_tpm, reference_information):
DeconRNASeq.load_r()
rpy2.robjects.r("data(liver_kidney)")
x = rpy2.robjects.r('signatures')
xdf = rpy2.robjects.pandas2ri.ri2py(x)
xdf.index = rpy2.robjects.r("rownames")(x)
xdf = xdf[xdf.kidney > xdf.liver]
return xdf # gg.convert_dataframe_to_r(x, True)
def load():
DeconRNASeq.load_r()
rpy2.robjects.r("data(liver_kidney)")
x = rpy2.robjects.r('datasets')
df = rpy2.robjects.pandas2ri.ri2py(x)
df.index = rpy2.robjects.r('rownames')(x)
sample_info = {}
for x in df.columns:
sample_info[x] = np.nan
return sample_info, df
def Percentage_DeconRNASeq():
DeconRNASeq.load_r()
def calc(df_tpm, genes_for_estimation):
if isinstance(genes_for_estimation, list):
signature_df = df_tpm.ix[genes_for_estimation, [
'reference_sample', 'reference_contamination']]
else:
signature_df = genes_for_estimation
if not isinstance(signature_df, rpy2.robjects.vectors.DataFrame):
signature_df = gg.convert_dataframe_to_r(signature_df, True)
if signature_df is None:
print genes_for_estimation
print type(genes_for_estimation)
raise ValueError("Here it is")
mixture_df = df_tpm[['observed']]
# for some arcane reason it won't work if you run it with one sample
# only...
mixture_df.insert(1, 'duplicate', mixture_df['observed'])
mixture_df = gg.convert_dataframe_to_r(mixture_df, True)
result = rpy2.robjects.r('DeconRNASeq')(
mixture_df,
signature_df,
)
calc_percentage = result[0][1] # the contamination percentage
calc_percentage_std = np.nan
percentage, corrected = correct_with_percentage(
df_tpm,
'observed',
['reference_sample'],
['reference_contamination'],
calc_percentage)
return {
'percentage': calc_percentage,
'percentage_std': calc_percentage_std,
'corrected': corrected,
}
return 'DeconRNAseq', calc
def select_genes(df_tpm, reference_information):
DeconRNASeq.load_r()
rpy2.robjects.r("data(rat_liver_brain)")
x = rpy2.robjects.r('array.signatures')
return list(rpy2.robjects.r('rownames')(x))
def select_genes(df_tpm, reference_information):
DeconRNASeq.load_r()
rpy2.robjects.r("data(liver_kidney)")
x = rpy2.robjects.r('signatures')
return x
