def get_save_isotope_patterns(db_filename, db_dump_folder, db_name, adduct, isocalc_sig, isocalc_resolution, charge, verbose, instrument):
import pickle
from pySM.parse_databases import parse_databases
# Check if already generated and load if possible, otherwise calculate fresh
sum_formulae = parse_databases.read_generic_csv(db_filename,header=0,idcol=0,namecol=1,sfcol=2, sep='\t')
load_file = '{}/{}_{}_{}_{}_{}.dbasedump'.format(db_dump_folder, db_name, instrument.__class__.__name__, adduct, isocalc_sig, isocalc_resolution)
if os.path.isfile(load_file):
if verbose:
print "{} -> loading".format(load_file)
try:
mz_list_tmp = pickle.load(open(load_file, 'r'))
except ValueError as e:
if verbose:
print str(e)
print "{} -> generating".format(load_file)
mz_list_tmp = calculate_isotope_patterns(sum_formulae, adduct=adduct, instrument=instrument, isocalc_sig=isocalc_sig,
isocalc_resolution=isocalc_resolution, charge=charge, verbose=False)
save_pattern(load_file, db_dump_folder, mz_list_tmp)
else:
if verbose:
print "{} -> generating".format(load_file)
mz_list_tmp = calculate_isotope_patterns(sum_formulae, adduct=adduct, isocalc_sig=isocalc_sig,
isocalc_resolution=isocalc_resolution, charge=charge, verbose=False, instrument=instrument)
save_pattern(load_file, db_dump_folder, mz_list_tmp)
return mz_list_tmp
def generate_isotope_patterns(config,verbose=True):
#todo: verbose -> logging.debug
from pySM.parse_databases import parse_databases
from pySM.spatial_metabolomics import get_save_isotope_patterns
from pyMSpec import instrument
# Extract variables from config dict
db_filenames = config['file_inputs']['database_file']
db_dump_folder = config['file_inputs']['database_load_folder']
#isocalc_sig = float(config['isotope_generation']['isocalc_sig'])
#isocalc_resolution = float(config['isotope_generation']['isocalc_resolution'])
if len(config['isotope_generation']['charge']) > 1:
print 'Warning: only first charge state currently accepted'
charge = int('{}{}'.format(config['isotope_generation']['charge'][0]['polarity'],
config['isotope_generation']['charge'][0][
'n_charges'])) # currently only supports first charge!!
adducts = set([a['adduct'] for a in config['isotope_generation']['adducts']])
instrument = get_instrument(config["isotope_generation"]["instrument"], config["isotope_generation"]["resolving_power"], config["isotope_generation"]["at_mz"])
isocalc_sig = np.round(instrument.sigma_at_mz(200), decimals=4)
isocalc_resolution = instrument.points_per_mz(isocalc_sig)
# Get master list of sum formulae
sum_formulae = {}
for db_filename in db_filenames:
sum_formulae = parse_databases.read_generic_csv(db_filename,header=0,idcol=0,namecol=1,sfcol=2, sep='\t', sum_formulae=sum_formulae)
if '' in sum_formulae:
if verbose:
print 'empty sf removed from list'
del sum_formulae['']
# Get isotope patterns for all sum_formulae
mz_list = {}
for db_filename in db_filenames:
db_name = os.path.splitext(os.path.basename(db_filename))[0]
for adduct in adducts:
_mz_list = get_save_isotope_patterns(db_filename, db_dump_folder, db_name, adduct, isocalc_sig, isocalc_resolution, charge, verbose, instrument)
# add patterns to total list
for sum_formula in sum_formulae:
if sum_formula not in _mz_list:# could be missing if [M-a] would have negative atoms
continue
if sum_formula not in mz_list:
mz_list[sum_formula]={}
## this limit of 4 is hardcoded to reduce the number of calculations todo: add to config file
n = np.min([4,len(_mz_list[sum_formula][adduct][0])])
sort_idx = np.argsort(_mz_list[sum_formula][adduct][1])[-n:][-1::-1]
mz_list[sum_formula][adduct] = [_mz_list[sum_formula][adduct][0][sort_idx],_mz_list[sum_formula][adduct][1][sort_idx]]
# Clean up
# poorly formatted formulae may not recieve an isotope pattern
rm_list = []
for sum_formula in sum_formulae:
if sum_formula not in mz_list:
rm_list.append(sum_formula)
if verbose:
print "{} formula to remove".format(len(rm_list))
for sum_formula in rm_list:
sum_formulae.pop(sum_formula,None)
if verbose:
print 'all isotope patterns generated and loaded'
return sum_formulae, adducts, mz_list
isocalc_resolution = instrument.points_per_mz(isocalc_sig)
db_dump_folder = config['file_inputs']['database_load_folder']
if len(config['isotope_generation']['charge']) > 1:
print 'Warning: only first charge state currently accepted'
charge = int('{}{}'.format(config['isotope_generation']['charge'][0]['polarity'],
config['isotope_generation']['charge'][0][
'n_charges'])) # currently only supports first charge!!
db_name = os.path.splitext(os.path.basename(db_filename))[0]
load_file = '{}/{}_{}_{}_{}_{}.dbasedump'.format(db_dump_folder, db_name, instrument.__class__.__name__, adduct, isocalc_sig, isocalc_resolution)
print "{} -> generating".format(load_file)
mz_list_tmp = calculate_isotope_patterns(sum_formulae, adduct=adduct, isocalc_sig=isocalc_sig,
isocalc_resolution=isocalc_resolution, charge=charge, verbose=False, instrument=instrument)
spatial_metabolomics.save_pattern(load_file, db_dump_folder, mz_list_tmp)
return True
if __name__== '__main__':
json_filename = "/home/palmer/Documents/tmp_data/MB/URenn/16135s1-3_mousebrain_dhbsub.json"
config = spatial_metabolomics.get_variables(json_filename)
adducts = [a['adduct'] for a in config['isotope_generation']['adducts']]
db_filenames = config['file_inputs']['database_file']
if isinstance(db_filenames, basestring):
db_filenames = [db_filenames,]
for db_filename in db_filenames:
sum_formulae = parse_databases.read_generic_csv(db_filename,header=1,idcol=0,namecol=1,sfcol=2, sep='\t')
if '' in sum_formulae:
del sum_formulae['']
for a in adducts:
generate_isotope_pattern(sum_formulae,a,config, db_filename)
adduct=adduct,
isocalc_sig=isocalc_sig,
isocalc_resolution=isocalc_resolution,
charge=charge,
verbose=False,
instrument=instrument)
spatial_metabolomics.save_pattern(load_file, db_dump_folder, mz_list_tmp)
return True
if __name__ == '__main__':
json_filename = "/home/palmer/Documents/tmp_data/MB/URenn/16135s1-3_mousebrain_dhbsub.json"
config = spatial_metabolomics.get_variables(json_filename)
adducts = [a['adduct'] for a in config['isotope_generation']['adducts']]
db_filenames = config['file_inputs']['database_file']
if isinstance(db_filenames, basestring):
db_filenames = [
db_filenames,
]
for db_filename in db_filenames:
sum_formulae = parse_databases.read_generic_csv(db_filename,
header=1,
idcol=0,
namecol=1,
sfcol=2,
sep='\t')
if '' in sum_formulae:
del sum_formulae['']
for a in adducts:
generate_isotope_pattern(sum_formulae, a, config, db_filename)