*Author:* Dénes Türei *Contact:* turei.denes@gmail.com *Git repositories (mirrored):* - https://git.embl.de/turei/lipyd- https://bitbucket.org/deeenes/lipyd- https://github.com/saezlab/lipydAuthor: Dénes Türei
Contact: turei.denes@gmail.com
Git repositories (mirrored):
import impimport itertoolsimport pprintfrom lipyd import pprint_namedtuplepprint.PrettyPrinter = pprint_namedtuple.PrettyPrinterfrom lipyd import massfrom lipyd import formulaThe `mass` module knows exact masses of isotopes, isotopic abundances, weights, etc. The `MassBase` class is able to process chemical formula, calculate masses and do arithmetics.The mass module knows exact masses of isotopes, isotopic abundances, weights, etc. The MassBase class is able to process chemical formula, calculate masses and do arithmetics.
mass.get_mass('Na')mass.MassBase('C2H6') - mass.MassBase('H') + mass.MassBase('OH')Expression evaluation with `mass.expr`:Expression evaluation with mass.expr:
mass.expr('C2H6 - H + OH')mass.expr('C6H12O6 - water')Make a deuterium:Make a deuterium:
mass.expr('H + n')Hydrogensulphate ion:Hydrogensulphate ion:
mass.expr('HSO4 + e')Additional attributes can be provided in keyword arguments to carry metadata.Additional attributes can be provided in keyword arguments to carry metadata.
lactic_acid = formula.Formula('CH3CHOHCOOH', name = 'lactic acid')lactic_acid.attrs.nameA galactose:A galactose:
((2 * formula.Formula('C6H12O6')) - 'H2O').formulafrom lipyd import mzThis is an oleic acid. What is the mass of the [M-H]- adduct?This is an oleic acid. What is the mass of the [M-H]- adduct?
formula.Formula('C18H34O2').remove_h()We have seen a mass and wondering what is the exact mass if it is an [M+NH4]+ adduct:We have seen a mass and wondering what is the exact mass if it is an [M+NH4]+ adduct:
mz.Mz(854.576482597791).remove_nh4()Calculate the [M+Li]+ adduct for the same molecule:Calculate the [M+Li]+ adduct for the same molecule:
mz.Mz(836.5426570444603).adduct(mass.MassBase('Li', charge = 1))Metabolites consist of a core and optionally substituents. Substituents might be formulas or moieties with aliphatic chains.Metabolites consist of a core and optionally substituents. Substituents might be formulas or moieties with aliphatic chains.
from lipyd import metabolitefrom lipyd import substituentMake all combinations of halogenated methanes:Make all combinations of halogenated methanes:
halo_methanes = metabolite.AbstractMetabolite( core = 'C', subs = [('H', 'F', 'Cl', 'Br', 'I')] * 4)Check the first 3:Check the first 3:
[(m.formula, m.mass) for m in halo_methanes][:3]Do the same with all alcohols up to 1-8 carbon count with 0-2 unsaturated bonds:Do the same with all alcohols up to 1-8 carbon count with 0-2 unsaturated bonds:
chain = metabolite.AbstractSubstituent(c = (1, 8), u = (0, 2))alcohols = metabolite.AbstractMetabolite(subs = (chain, ('OH',)))[(m.formula, m.mass) for m in alcohols][:3]Make some ceramides:Make some ceramides:
# fatty acyls of length 16, 18 or 20 and one or no unsaturation:fattyacyl = substituent.FattyAcyl(c = (16, 18, 20), u = (0, 1))lcb = substituent.Sphingosine(c = 18, u = 1)ceramides = metabolite.AbstractMetabolite(core = 'H', subs = (lcb, fattyacyl), name = 'Cer')# name, formula, mass, [M+H]+, [M+NH4]+[(cer.name, cer.formula, cer.mass, cer.add_nh4()) for cer in ceramides]In the `lipyd.lipid` module more than 150 lipid varieties are predefined.In the lipyd.lipid module more than 150 lipid varieties are predefined.
from lipyd import lipidd_ceramides = lipid.CeramideD( sph_args = {'c': 18, 'u': (0, 1)}, fa_args = {'c': (14, 24), 'u': (0, 4), 'even': True},)[(cer.name, cer.mass) for cer in d_ceramides][:3]The `lipyd.moldb` module provides access to *SwissLipis* and *LipidMaps*. Automatically downloads and processes the databases which you can search by various names and identifiers, you can also access structures as OpenBabel objects, InChI and SMILE strings.The lipyd.moldb module provides access to SwissLipis and LipidMaps. Automatically downloads and processes the databases which you can search by various names and identifiers, you can also access structures as OpenBabel objects, InChI and SMILE strings.
from lipyd import moldbfrom lipyd import namefrom lipyd.lipproc import *swl = moldb.SwissLipids(levels = {'species'})Get phosphatidylethanolamines as OpenBabel objects:Get phosphatidylethanolamines as OpenBabel objects:
swl.reload()pe = swl.get_hg_obmol('PE')pe0 = next(pe)pe0.draw()[m.title for m in itertools.islice(pe, 0, 3)]lm = moldb.LipidMaps()gibberellin = list(lm.get_record('LMPR0104170034', typ = 'mainkey'))[0]gibberellin['name']LipidMaps too is able to yield OpenBabel objects:LipidMaps too is able to yield OpenBabel objects:
tag = list(lm.get_obmol('TAG(15:0_20:4_20:5)', 'synonym'))[0]tag.draw()In order to make the databases computationally useful and use them as a combined database, we need to process their nomenclature. The `lipyd.name` module is able to recognize dozens of lipid names used in SwissLipids and LipidMaps.In order to make the databases computationally useful and use them as a combined database, we need to process their nomenclature. The lipyd.name module is able to recognize dozens of lipid names used in SwissLipids and LipidMaps.
from lipyd import namenameproc = name.LipidNameProcessor(database = 'swisslipids', iso = True)processed_name = nameproc.process( ['Phosphatidylethanolamine (16:0/20:4(5Z,8Z,11Z,14Z))'])pprint.pprint(processed_name)It understands even greek names:It understands even greek names:
nameproc.process(['eicosapentaenoate'])db = moldb.MoleculeDatabaseAggregator()Either exact masses or adducts can be searched in the database by `lookup` and `adduct_lookup` methods, respectively.Either exact masses or adducts can be searched in the database by lookup and adduct_lookup methods, respectively.
result = db.adduct_lookup(757.549011, ionmode = 'pos')Take a closer look at one of the resulted records:Take a closer look at one of the resulted records:
pprint.pprint(result['[M+NH4]+'][1][2])The exact masses and errors for all hits are also provided. Errors in ppm:The exact masses and errors for all hits are also provided. Errors in ppm:
result['[M+NH4]+'][2]Repeat the lookup with lower tolerance, and the items with high ppm disappear:Repeat the lookup with lower tolerance, and the items with high ppm disappear:
result = db.adduct_lookup(757.549011, ionmode = 'pos', tolerance = 5)result['[M+NH4]+'][2]The fragment database provided by `lipyd.fragment` and `lipyd.fragdb` modules works similar way as `lipyd.lipid` and `lipyd.moldb`. `lipyd.fragment` contains near 100 predefined aliphatic chain derived fragments. In addition 140 headgroup derived fragments are included like for example 184 for choline.The fragment database provided by lipyd.fragment and lipyd.fragdb modules works similar way as lipyd.lipid and lipyd.moldb. lipyd.fragment contains near 100 predefined aliphatic chain derived fragments. In addition 140 headgroup derived fragments are included like for example 184 for choline.
from lipyd import fragmentfrom lipyd import fragdbAs an example take a look at a [Sph-NH2-OH]- fragment:As an example take a look at a [Sph-NH2-OH]- fragment:
sphfrag = fragment.Sph_mNH2_mOH(c = 18, u = 1)At this fragment type the constraints tell us which lipid varieties this fragment can be observed. In this case *dCer* and *DHCer*.At this fragment type the constraints tell us which lipid varieties this fragment can be observed. In this case dCer and DHCer.
sphfrag.constraintslist(sphfrag)[0].charge, list(sphfrag)[0].massLook up a negative mode fragment m/z in the database. It results an array with mass, fragment name, fragment type, aliphatic chain type, carbon count, unsaturation and charge in each row. At neutral losses the charge is 0.Look up a negative mode fragment m/z in the database. It results an array with mass, fragment name, fragment type, aliphatic chain type, carbon count, unsaturation and charge in each row. At neutral losses the charge is 0.
fragdb.lookup_neg(283.26)Now let's annotate an MS2 scan with possible fragment identifications. To do this we open an example MGF file included in the module. The `lipyd.mgf` module serves MS2 scans from MGF files on demand. Btw the `lipyd.settings` module gives easy access for and control over near 100 customizable parameters.Now let's annotate an MS2 scan with possible fragment identifications. To do this we open an example MGF file included in the module. The lipyd.mgf module serves MS2 scans from MGF files on demand. Btw the lipyd.settings module gives easy access for and control over near 100 customizable parameters.
from lipyd import mgffrom lipyd import settingsmgffile = settings.get('mgf_example')mgfreader = mgf.MgfReader(mgffile)precursor = 590.45536 # this is a Cer-1Pidx, rtdiff = mgfreader.lookup_scan_ids(precursor)We found the following scans for precursor 590.455:We found the following scans for precursor 590.455:
idxSelect a scan from the ones above and annotate its fragments:Select a scan from the ones above and annotate its fragments:
scan = mgfreader.scan_by_id(1941)annot = fragdb.FragmentAnnotator( mzs = scan[:,0], ionmode = 'pos', precursor = precursor)One example of the annotations. This fragment ranks 25 by intensity.One example of the annotations. This fragment ranks 25 by intensity.
pprint.pprint(list(annot)[24])The `lipyd.ms2.Scan` class is able to perform the entire identification workflow. By an alternative constructor method it can be initialized by providing and MGF file and scan ID.The lipyd.ms2.Scan class is able to perform the entire identification workflow. By an alternative constructor method it can be initialized by providing and MGF file and scan ID.
from lipyd import ms2mgfname = settings.get('mgf_pos_examples')scan_id = 3626scan = ms2.Scan.from_mgf(mgfname, scan_id, 'pos')If not provided the `Scan` instance performs the database lookup of the precursor ion. Here are the results:If not provided the Scan instance performs the database lookup of the precursor ion. Here are the results:
pprint.pprint(scan.ms1_records['[M+H]+'][1][0])The `identify` method attempts to confirm each of the records by analysing the MS2 spectrum.The identify method attempts to confirm each of the records by analysing the MS2 spectrum.
identity = scan.identify()The results are grouped by lipid species and come with a score. Hex2-Cer(t42:2) got a score of 45, which is the highest at this scan:The results are grouped by lipid species and come with a score. Hex2-Cer(t42:2) got a score of 45, which is the highest at this scan:
pprint.pprint(identity['Hex2-Cer(t42:2)'][0])At the same time there were attempts to confirm for example Hex-Cer(d53:9-2OH) but it resulted a score of 0.At the same time there were attempts to confirm for example Hex-Cer(d53:9-2OH) but it resulted a score of 0.
pprint.pprint(identity['Hex-Cer(d53:9-2OH)'][0])Let's see one more example.Let's see one more example.
mgfname = settings.get('mgf_neg_examples')scan_id = 2516scan = ms2.Scan.from_mgf(mgfname, scan_id, 'neg')identity = scan.identify()We see that this is a PI(34:1) with score 11 and both fatty acyl fragments are confirmed by [FA-H]- ions (see the `ChainIdentificationDetails` object). These fragments are the 1st and 2nd most abundant with relative intensities of 100% and 99%.We see that this is a PI(34:1) with score 11 and both fatty acyl fragments are confirmed by [FA-H]- ions (see the ChainIdentificationDetails object). These fragments are the 1st and 2nd most abundant with relative intensities of 100% and 99%.
pprint.pprint(identity['PI(34:1)'][0])