# -*- coding: utf-8 -*-
"""
Functions to read and process data generated with the Kojak cross-link
search engine.
"""
import numpy as np
import pandas as pd
if __name__ == '__main__':
import HelperFunctions as hf
import KojakFunctions as kj
else:
from . import HelperFunctions as hf
from . import KojakFunctions as kj
[docs]def Read(kojak_files, rawfile=None, decoy_string='decoy', col_order=None, compact=False):
"""
Read Kojak results file, calculate and process missing values required
for xTable and return the xTable.
Args:
kojak_files (list): path or paths to Kojak results file(s)
rawfile (str): name of the corresponding rawfile
decoy_string (optional): string used in kojak to label decoys
col_order (list) – List of xTable column titles that are used to sort and compress the resulting datatable
compact (bool): Compact the xTable to only the columns given in col_order or not
Returns:
pandas.DataFrame: xtable data table
"""
# convert to list if the input is only a single path
if not isinstance(kojak_files, list):
kojak_files = [kojak_files]
allData = list()
kojak_dtypes = {'Scan Number': pd.Int64Dtype(),
'Charge': pd.Int64Dtype(),
'Link #1': pd.Int64Dtype(),
'Link #2': pd.Int64Dtype(),
'Score': float
}
for file in kojak_files:
print('Reading Kojak-file: ' + file)
# only called if kojak_file is not None
try:
s = pd.read_csv(hf.compatible_path(file),
skiprows = 1, # skip the Kojak version
dtype=kojak_dtypes,
na_values = '-',
delimiter='\t')
allData.append(s)
except Exception as e:
raise Exception('[xTable Read] Failed opening file: {}'.format(file))
xtable = pd.concat(allData)
# remove lines containing non-identified PSMs (marked with '-' in both
# Link columns
xtable.dropna(axis=0, how='all', subset=['Link #1', 'Link #2'], inplace=True)
# dropping lines causes fragmented index --> regenate the index
xtable.reset_index(drop=True, inplace=True)
# if split into mulitple rows if multiple candidate proteins were found to
# match an experimental spectrum
xtable = hf.split_concatenated_lists(xtable, where=['Protein #1', 'Protein #2'])
### Process the data to comply to xTable format
xtable = xtable.rename(columns={'Scan Number': 'scanno',
'Charge': 'prec_ch',
'Link #1': 'xlink1',
'Link #2': 'xlink2',
'Score': 'score'
})
# Extract peptide sequence, modification mass and position from the
# Peptide #1 and Peptide #2 entries
xtable = kj.extract_peptide(xtable)
# transform unset xlinks to np.nan
xtable[['xlink1', 'xlink2']] = xtable[['xlink1', 'xlink2']].replace(-1, np.nan)
# extract protein name and relative cross-link position from the Protein #
# entries
xtable = kj.extract_protein(xtable)
# calculate absolute position of first AA of peptide
# ignoring errors avoids raising error in case on NaN -> returns NaN
# as pos
# Must be calculated as float as NaN is not implemented in int
xtable['pos1'] =\
xtable['xpos1'].astype(float, errors='ignore') - \
xtable['xlink1'].astype(float, errors='ignore') + 1
xtable['pos2'] =\
xtable['xpos2'].astype(float, errors='ignore') - \
xtable['xlink2'].astype(float, errors='ignore') + 1
# Calculate if a cross link is of inter or of loop type
# Refine the inter type into inter/intra/homomultimeric
# Generate ID for the xlinks
xtable = kj.assign_ID_and_type(xtable)
#sets the column decoy based on whether the decoy string is present in the
# protein name or not
xtable = kj.set_decoy(xtable, decoy_string)
# set the rawfile name for xtable (None if not provided by call)
xtable['rawfile'] = rawfile
xtable['xtype'] = np.nan
xtable['search_engine'] = 'Kojak'
xtable = hf.order_columns(xtable, col_order, compact)
### return xtable df
return xtable
if __name__ == '__main__':
kojak_file = r'C:\Users\User\Documents\03_software\python\CroCo\testdata\final\input\kojak\20180615_KS_CL_9_msconvert.kojak.txt'
col_order = ['rawfile', 'scanno', 'prec_ch',
'pepseq1', 'xlink1',
'pepseq2', 'xlink2', 'xtype',
'modmass1', 'modpos1', 'mod1',
'modmass2', 'modpos2', 'mod2',
'prot1', 'xpos1', 'prot2',
'xpos2', 'type', 'score', 'ID', 'pos1', 'pos2', 'decoy']
xtable = Read(kojak_file, col_order=col_order, rawfile='20180615_KS_CL_9')
