def melt_df(df1):
"""
This function melts the dataframe in long form based on some fixed columns
Args:
df : dataframe to be converted in long format
Returns:
long_form : dataframe in long format
"""
fixed_cols = [maven_constants.NAME, maven_constants.LABEL, maven_constants.FORMULA]
col_headers = df1.columns.tolist()
check_column_headers(col_headers, fixed_cols)
melt_cols = [x for x in col_headers if x not in fixed_cols]
# try:
long_form = pd.melt(df1, id_vars=fixed_cols, value_vars=melt_cols)
# except KeyError():
# raise KeyError('columns {} not found in input data'.format(','.join(fixed_cols)))
return long_form
def merge_samples(merged_df, sample_metadata):
"""
This function merge the raw input dataframe with the sample metadata
dataframe.
"""
if sample_metadata is not None:
col_headers_sample = sample_metadata.columns.values
col_headers_merged = merged_df.columns.values
bg_corr_col_names_sample = [
multiquant.BACKGROUND, multiquant.MQ_COHORT_NAME
]
bg_corr_col_names_merged = [multiquant.MQ_COHORT_NAME]
try:
hlp.check_column_headers(col_headers_sample,
bg_corr_col_names_sample)
hlp.check_column_headers(col_headers_merged,
bg_corr_col_names_merged)
assert set(sample_metadata[multiquant.BACKGROUND]).issubset(
set(sample_metadata[multiquant.MQ_SAMPLE_NAME]))
merged_df = merged_df.merge(
sample_metadata,
how='inner',
on=[multiquant.MQ_SAMPLE_NAME, multiquant.MQ_COHORT_NAME])
if merged_df.empty:
raise Exception(
'Empty Merge, no common entries to process, please check input files'
)
except AssertionError:
warnings.warn("Background Correction can't be performed")
merged_df = merged_df.merge(sample_metadata,
how='inner',
on=[multiquant.MQ_SAMPLE_NAME])
# first change Sample Name to Cohort Name, then Original Filename to Sample
# refer to multiquant raw output
merged_df.rename(columns={multiquant.MQ_COHORT_NAME: multiquant.COHORT},
inplace=True)
merged_df.rename(columns={multiquant.MQ_SAMPLE_NAME: multiquant.SAMPLE},
inplace=True)
remove_stds = remove_mq_stds(merged_df)
remove_stds.rename(columns={"Area": multiquant.INTENSITY}, inplace=True)
return remove_stds
def merge_mq_metadata(mq_df, metdata, sample_metdata):
"""
This function merge the raw input dataframe with the metadata and
sample metadata.
Args:
mq_df: raw input df
metdata: metadata df
sample_metdata: sample metadata df
Returns:
merged_df: merged dataframe
list_of_replicates: list of sample names belonging to same cohort.
"""
merged_data = mq_merge_dfs(mq_df, metdata, sample_metdata)
merged_data.fillna(0, inplace=True)
list_of_replicates = []
sample_background = []
if sample_metdata is not None:
col_headers = merged_data.columns.values
bg_corr_col_names = [multiquant.BACKGROUND, multiquant.COHORT]
try:
hlp.check_column_headers(col_headers, bg_corr_col_names)
except AssertionError:
return merged_data, list_of_replicates, sample_background
# consider only those sample names which are present in raw file
sample_metdata = sample_metdata[sample_metdata[
multiquant.MQ_SAMPLE_NAME].isin(merged_data[multiquant.SAMPLE])]
list_of_replicates = get_replicates(sample_metdata,
multiquant.MQ_SAMPLE_NAME,
multiquant.MQ_COHORT_NAME,
multiquant.BACKGROUND)
sample_background = get_background_samples(sample_metdata,
multiquant.MQ_SAMPLE_NAME,
multiquant.BACKGROUND)
return merged_data, list_of_replicates, sample_background