Esempio n. 1
0
def prody_pca(coords, **kwargs):
    """Perform PCA calculations for PDB or DCD format *coords* file.

    """

    for key in DEFAULTS:
        if not key in kwargs:
            kwargs[key] = DEFAULTS[key]

    from os.path import isdir, splitext, join
    outdir = kwargs.get('outdir')
    if not isdir(outdir):
        raise IOError('{0} is not a valid path'.format(repr(outdir)))

    import prody
    LOGGER = prody.LOGGER

    prefix = kwargs.get('prefix')
    nmodes = kwargs.get('nmodes')
    selstr = kwargs.get('select')

    ext = splitext(coords)[1].lower()
    if ext == '.gz':
        ext = splitext(coords[:-3])[1].lower()

    if ext == '.dcd':
        pdb = kwargs.get('psf') or kwargs.get('pdb')
        if pdb:
            if splitext(pdb)[1].lower() == '.psf':
                pdb = prody.parsePSF(pdb)
            else:
                pdb = prody.parsePDB(pdb)
        dcd = prody.DCDFile(coords)
        if prefix == '_pca' or prefix == '_eda':
            prefix = dcd.getTitle() + prefix

        if len(dcd) < 2:
            raise ValueError('DCD file must have multiple frames')
        if pdb:
            if pdb.numAtoms() == dcd.numAtoms():
                select = pdb.select(selstr)
                dcd.setAtoms(select)
                LOGGER.info('{0} atoms are selected for calculations.'
                            .format(len(select)))
            else:
                select = pdb.select(selstr)
                if select.numAtoms() != dcd.numAtoms():
                    raise ValueError('number of selected atoms ({0}) does '
                                     'not match number of atoms in the DCD '
                                     'file ({1})'.format(select.numAtoms(),
                                                           dcd.numAtoms()))
                if pdb.numCoordsets():
                    dcd.setCoords(select.getCoords())

        else:
            select = prody.AtomGroup()
            select.setCoords(dcd.getCoords())
        pca = prody.PCA(dcd.getTitle())
        if len(dcd) > 1000:
            pca.buildCovariance(dcd, aligned=kwargs.get('aligned'))
            pca.calcModes(nmodes)
            ensemble = dcd
        else:
            ensemble = dcd[:]
            if not kwargs.get('aligned'):
                ensemble.iterpose()
            pca.performSVD(ensemble)

    else:
        pdb = prody.parsePDB(coords)
        if pdb.numCoordsets() < 2:
            raise ValueError('PDB file must contain multiple models')

        if prefix == '_pca' or prefix == '_eda':
            prefix = pdb.getTitle() + prefix

        select = pdb.select(selstr)
        LOGGER.info('{0} atoms are selected for calculations.'
                    .format(len(select)))
        if select is None:
            raise ValueError('selection {0} do not match any atoms'
                                .format(repr(selstr)))
        LOGGER.info('{0} atoms will be used for PCA calculations.'
                    .format(len(select)))
        ensemble = prody.Ensemble(select)
        pca = prody.PCA(pdb.getTitle())
        if not kwargs.get('aligned'):
            ensemble.iterpose()
        pca.performSVD(ensemble)


    LOGGER.info('Writing numerical output.')
    if kwargs.get('outnpz'):
        prody.saveModel(pca, join(outdir, prefix))

    prody.writeNMD(join(outdir, prefix + '.nmd'), pca[:nmodes], select)

    extend = kwargs.get('extend')
    if extend:
        if pdb:
            if extend == 'all':
                extended = prody.extendModel(pca[:nmodes], select, pdb)
            else:
                extended = prody.extendModel(pca[:nmodes], select,
                                             select | pdb.bb)
            prody.writeNMD(join(outdir, prefix + '_extended_' +
                           extend + '.nmd'), *extended)
        else:
            prody.LOGGER.warn('Model could not be extended, provide a PDB or '
                              'PSF file.')
    outall = kwargs.get('outall')
    delim = kwargs.get('numdelim')
    ext = kwargs.get('numext')
    format = kwargs.get('numformat')

    if outall or kwargs.get('outeig'):
        prody.writeArray(join(outdir, prefix + '_evectors'+ext),
                         pca.getArray(), delimiter=delim, format=format)
        prody.writeArray(join(outdir, prefix + '_evalues'+ext),
                         pca.getEigvals(), delimiter=delim, format=format)
    if outall or kwargs.get('outcov'):
        prody.writeArray(join(outdir, prefix + '_covariance'+ext),
                         pca.getCovariance(), delimiter=delim, format=format)
    if outall or kwargs.get('outcc') or kwargs.get('outhm'):
        cc = prody.calcCrossCorr(pca)
        if outall or kwargs.get('outcc'):
            prody.writeArray(join(outdir, prefix + '_cross-correlations' +
                             ext), cc, delimiter=delim, format=format)
        if outall or kwargs.get('outhm'):
            resnums = select.getResnums()
            hmargs = {} if resnums is None else {'resnums': resnums}
            prody.writeHeatmap(join(outdir, prefix + '_cross-correlations.hm'),
                               cc, xlabel='Residue', ylabel='Residue',
                               title=pca.getTitle() + ' cross-correlations',
                               **hmargs)

    if outall or kwargs.get('outsf'):
        prody.writeArray(join(outdir, prefix + '_sqfluct'+ext),
                         prody.calcSqFlucts(pca), delimiter=delim,
                         format=format)
    if outall or kwargs.get('outproj'):
        prody.writeArray(join(outdir, prefix + '_proj'+ext),
                         prody.calcProjection(ensemble, pca), delimiter=delim,
                         format=format)

    figall = kwargs.get('figall')
    cc = kwargs.get('figcc')
    sf = kwargs.get('figsf')
    sp = kwargs.get('figproj')

    if figall or cc or sf or sp:
        try:
            import matplotlib.pyplot as plt
        except ImportError:
            LOGGER.warning('Matplotlib could not be imported. '
                           'Figures are not saved.')
        else:
            prody.SETTINGS['auto_show'] = False
            LOGGER.info('Saving graphical output.')
            format = kwargs.get('figformat')
            width = kwargs.get('figwidth')
            height = kwargs.get('figheight')
            dpi = kwargs.get('figdpi')

            format = format.lower()
            if figall or cc:
                plt.figure(figsize=(width, height))
                prody.showCrossCorr(pca)
                plt.savefig(join(outdir, prefix + '_cc.'+format),
                    dpi=dpi, format=format)
                plt.close('all')
            if figall or sf:
                plt.figure(figsize=(width, height))
                prody.showSqFlucts(pca)
                plt.savefig(join(outdir, prefix + '_sf.'+format),
                    dpi=dpi, format=format)
                plt.close('all')
            if figall or sp:
                indices = []
                for item in sp.split():
                    try:
                        if '-' in item:
                            item = item.split('-')
                            if len(item) == 2:
                                indices.append(list(range(int(item[0])-1,
                                                          int(item[1]))))
                        elif ',' in item:
                            indices.append([int(i)-1 for i in item.split(',')])
                        else:
                            indices.append(int(item)-1)
                    except:
                        pass
                for index in indices:
                        plt.figure(figsize=(width, height))
                        prody.showProjection(ensemble, pca[index])
                        if isinstance(index, int):
                            index = [index]
                        index = [str(i+1) for i in index]
                        plt.savefig(join(outdir, prefix + '_proj_' +
                            '_'.join(index) + '.' + format),
                            dpi=dpi, format=format)
                        plt.close('all')
Esempio n. 2
0
def prody_pca(opt):
    """Perform PCA calculations based on command line arguments."""
    
    outdir = opt.outdir
    if not os.path.isdir(outdir):
        opt.subparser.error('{0:s} is not a valid path'.format(outdir))
        
    import prody
    LOGGER = prody.LOGGER
        
    coords = opt.coords
    prefix = opt.prefix
    nmodes, selstr = opt.nmodes, opt.select
    
    if os.path.splitext(coords)[1].lower() == '.dcd':     
        ag = opt.psf or opt.pdb
        if ag:
            if os.path.splitext(ag)[1].lower() == '.psf':
                ag = prody.parsePSF(ag)
            else:
                ag = prody.parsePDB(ag)
        dcd = prody.DCDFile(opt.coords)
        if len(dcd) < 2:
            opt.subparser("DCD file must contain multiple frames.")
        if ag:
            dcd.setAtomGroup(ag)
            select = dcd.select(selstr)
            LOGGER.info('{0:d} atoms are selected for calculations.'
                        .format(len(select)))
        else:
            select = prody.AtomGroup()
            select.setCoords(dcd.getCoords())
        pca = prody.PCA(dcd.getTitle())
        if len(dcd) > 1000:
            pca.buildCovariance(dcd)
            pca.calcModes(dcd)
        else:
            pca.performSVD(dcd[:])
    else:
        pdb = prody.parsePDB(opt.coords)
        if pdb.numCoordsets() < 2:
            opt.subparser("PDB file must contain multiple models.")
        if prefix == '_pca':
            prefix = pdb.getTitle() + '_pca'
        select = pdb.select(selstr)
        LOGGER.info('{0:d} atoms are selected for calculations.'
                    .format(len(select)))
        if select is None:
            opt.subparser('Selection "{0:s}" do not match any atoms.'
                          .format(selstr))
        LOGGER.info('{0:d} atoms will be used for PCA calculations.'
                    .format(len(select)))
        ensemble = prody.Ensemble(select)
        pca = prody.PCA(pdb.getTitle())
        ensemble.iterpose()
        pca.performSVD(ensemble)

    LOGGER.info('Writing numerical output.')
    if opt.npz:
        prody.saveModel(pca)
    prody.writeNMD(os.path.join(outdir, prefix + '.nmd'), pca[:nmodes], select)

    outall = opt.all
    delim, ext, format = opt.delim, opt.ext, opt.numformat
    if outall or opt.eigen:
        prody.writeArray(os.path.join(outdir, prefix + '_evectors'+ext), 
                         pca.getArray(), delimiter=delim, format=format)
        prody.writeArray(os.path.join(outdir, prefix + '_evalues'+ext), 
                         pca.getEigenvalues(), delimiter=delim, format=format)
    if outall or opt.covar:
        prody.writeArray(os.path.join(outdir, prefix + '_covariance'+ext), 
                         pca.getCovariance(), delimiter=delim, format=format)
    if outall or opt.ccorr:
        prody.writeArray(os.path.join(outdir, prefix + '_cross-correlations' + 
                                              ext), prody.calcCrossCorr(pca), 
                         delimiter=delim, format=format)
    if outall or opt.sqflucts:
        prody.writeArray(os.path.join(outdir, prefix + '_sqfluct'+ext), 
                         prody.calcSqFlucts(pca), delimiter=delim, 
                         format=format)
    if outall or opt.proj:
        prody.writeArray(os.path.join(outdir, prefix + '_proj'+ext), 
                         prody.calcProjection(ensemble, pca), delimiter=delim, 
                         format=format)
          
    figall, cc, sf, sp = opt.figures, opt.cc, opt.sf, opt.sp

    if figall or cc or sf or sp: 
        try:
            import matplotlib.pyplot as plt
        except ImportError:
            LOGGER.warning('Matplotlib could not be imported. '
                           'Figures are not saved.')
        else:
            LOGGER.info('Saving graphical output.')
            format, width, height, dpi = \
                opt.figformat, opt.width, opt.height, opt.dpi
            format = format.lower()
            if figall or cc:
                plt.figure(figsize=(width, height))
                prody.showCrossCorr(pca)
                plt.savefig(os.path.join(outdir, prefix + '_cc.'+format), 
                    dpi=dpi, format=format)
                plt.close('all')
            if figall or sf:
                plt.figure(figsize=(width, height))
                prody.showSqFlucts(pca)
                plt.savefig(os.path.join(outdir, prefix + '_sf.'+format), 
                    dpi=dpi, format=format)
                plt.close('all')                    
            if figall or sp:
                indices = []
                for item in sp.split():
                    try:
                        if '-' in item:
                            item = item.split('-')
                            if len(item) == 2:
                                indices.append(range(int(item[0])-1, 
                                               int(item[1])))
                        elif ',' in item:
                            indices.append([int(i)-1 for i in item.split(',')])
                        else:
                            indices.append(int(item)-1)
                    except:
                        pass
                for index in indices:
                        plt.figure(figsize=(width, height))
                        prody.showProjection(ensemble, pca[index])
                        if isinstance(index, int):
                            index = [index]
                        index = [str(i+1) for i in index]
                        plt.savefig(os.path.join(outdir, prefix + '_proj_' + 
                            '_'.join(index) + '.' + format),
                            dpi=dpi, format=format)
                        plt.close('all')                  
Esempio n. 3
0
def prody_pca(coords, **kwargs):
    """Perform PCA calculations for PDB or DCD format *coords* file.

    """

    for key in DEFAULTS:
        if not key in kwargs:
            kwargs[key] = DEFAULTS[key]

    from os.path import isdir, splitext, join
    outdir = kwargs.get('outdir')
    if not isdir(outdir):
        raise IOError('{0} is not a valid path'.format(repr(outdir)))

    import prody
    LOGGER = prody.LOGGER

    prefix = kwargs.get('prefix')
    nmodes = kwargs.get('nmodes')
    selstr = kwargs.get('select')
    quiet = kwargs.pop('quiet', False)
    altloc = kwargs.get('altloc')

    ext = splitext(coords)[1].lower()
    if ext == '.gz':
        ext = splitext(coords[:-3])[1].lower()

    if ext == '.dcd':
        pdb = kwargs.get('psf') or kwargs.get('pdb')
        if pdb:
            if splitext(pdb)[1].lower() == '.psf':
                pdb = prody.parsePSF(pdb)
            else:
                pdb = prody.parsePDB(pdb, altlocs=altlocs)
        dcd = prody.DCDFile(coords)
        if prefix == '_pca' or prefix == '_eda':
            prefix = dcd.getTitle() + prefix

        if len(dcd) < 2:
            raise ValueError('DCD file must have multiple frames')
        if pdb:
            if pdb.numAtoms() == dcd.numAtoms():
                select = pdb.select(selstr)
                dcd.setAtoms(select)
                LOGGER.info('{0} atoms are selected for calculations.'.format(
                    len(select)))
            else:
                select = pdb.select(selstr)
                if select.numAtoms() != dcd.numAtoms():
                    raise ValueError('number of selected atoms ({0}) does '
                                     'not match number of atoms in the DCD '
                                     'file ({1})'.format(
                                         select.numAtoms(), dcd.numAtoms()))
                if pdb.numCoordsets():
                    dcd.setCoords(select.getCoords())

        else:
            select = prody.AtomGroup()
            select.setCoords(dcd.getCoords())
        pca = prody.PCA(dcd.getTitle())

        nproc = kwargs.get('nproc')
        if nproc:
            try:
                from threadpoolctl import threadpool_limits
            except ImportError:
                raise ImportError(
                    'Please install threadpoolctl to control threads')

            with threadpool_limits(limits=nproc, user_api="blas"):
                if len(dcd) > 1000:
                    pca.buildCovariance(dcd,
                                        aligned=kwargs.get('aligned'),
                                        quiet=quiet)
                    pca.calcModes(nmodes)
                    ensemble = dcd
                else:
                    ensemble = dcd[:]
                    if not kwargs.get('aligned'):
                        ensemble.iterpose(quiet=quiet)
                    pca.performSVD(ensemble)
                nmodes = pca.numModes()
        else:
            if len(dcd) > 1000:
                pca.buildCovariance(dcd,
                                    aligned=kwargs.get('aligned'),
                                    quiet=quiet)
                pca.calcModes(nmodes)
                ensemble = dcd
            else:
                ensemble = dcd[:]
                if not kwargs.get('aligned'):
                    ensemble.iterpose(quiet=quiet)
                pca.performSVD(ensemble)
            nmodes = pca.numModes()

    else:
        pdb = prody.parsePDB(coords)
        if pdb.numCoordsets() < 2:
            raise ValueError('PDB file must contain multiple models')

        if prefix == '_pca' or prefix == '_eda':
            prefix = pdb.getTitle() + prefix

        select = pdb.select(selstr)
        LOGGER.info('{0} atoms are selected for calculations.'.format(
            len(select)))
        if select is None:
            raise ValueError('selection {0} do not match any atoms'.format(
                repr(selstr)))
        LOGGER.info('{0} atoms will be used for PCA calculations.'.format(
            len(select)))
        ensemble = prody.Ensemble(select)
        pca = prody.PCA(pdb.getTitle())
        if not kwargs.get('aligned'):
            ensemble.iterpose()

        nproc = kwargs.get('nproc')
        if nproc:
            try:
                from threadpoolctl import threadpool_limits
            except ImportError:
                raise ImportError(
                    'Please install threadpoolctl to control threads')

            with threadpool_limits(limits=nproc, user_api="blas"):
                pca.performSVD(ensemble)
        else:
            pca.performSVD(ensemble)

    LOGGER.info('Writing numerical output.')
    if kwargs.get('outnpz'):
        prody.saveModel(pca, join(outdir, prefix))

    if kwargs.get('outscipion'):
        prody.writeScipionModes(outdir, pca)

    prody.writeNMD(join(outdir, prefix + '.nmd'), pca[:nmodes], select)

    extend = kwargs.get('extend')
    if extend:
        if pdb:
            if extend == 'all':
                extended = prody.extendModel(pca[:nmodes], select, pdb)
            else:
                extended = prody.extendModel(pca[:nmodes], select,
                                             select | pdb.bb)
            prody.writeNMD(
                join(outdir, prefix + '_extended_' + extend + '.nmd'),
                *extended)
        else:
            prody.LOGGER.warn('Model could not be extended, provide a PDB or '
                              'PSF file.')
    outall = kwargs.get('outall')
    delim = kwargs.get('numdelim')
    ext = kwargs.get('numext')
    format = kwargs.get('numformat')

    if outall or kwargs.get('outeig'):
        prody.writeArray(join(outdir, prefix + '_evectors' + ext),
                         pca.getArray(),
                         delimiter=delim,
                         format=format)
        prody.writeArray(join(outdir, prefix + '_evalues' + ext),
                         pca.getEigvals(),
                         delimiter=delim,
                         format=format)
    if outall or kwargs.get('outcov'):
        prody.writeArray(join(outdir, prefix + '_covariance' + ext),
                         pca.getCovariance(),
                         delimiter=delim,
                         format=format)
    if outall or kwargs.get('outcc') or kwargs.get('outhm'):
        cc = prody.calcCrossCorr(pca)
        if outall or kwargs.get('outcc'):
            prody.writeArray(join(outdir,
                                  prefix + '_cross-correlations' + ext),
                             cc,
                             delimiter=delim,
                             format=format)
        if outall or kwargs.get('outhm'):
            resnums = select.getResnums()
            hmargs = {} if resnums is None else {'resnums': resnums}
            prody.writeHeatmap(join(outdir, prefix + '_cross-correlations.hm'),
                               cc,
                               xlabel='Residue',
                               ylabel='Residue',
                               title=pca.getTitle() + ' cross-correlations',
                               **hmargs)

    if outall or kwargs.get('outsf'):
        prody.writeArray(join(outdir, prefix + '_sqfluct' + ext),
                         prody.calcSqFlucts(pca),
                         delimiter=delim,
                         format=format)
    if outall or kwargs.get('outproj'):
        prody.writeArray(join(outdir, prefix + '_proj' + ext),
                         prody.calcProjection(ensemble, pca),
                         delimiter=delim,
                         format=format)

    figall = kwargs.get('figall')
    cc = kwargs.get('figcc')
    sf = kwargs.get('figsf')
    sp = kwargs.get('figproj')

    if figall or cc or sf or sp:
        try:
            import matplotlib.pyplot as plt
        except ImportError:
            LOGGER.warning('Matplotlib could not be imported. '
                           'Figures are not saved.')
        else:
            prody.SETTINGS['auto_show'] = False
            LOGGER.info('Saving graphical output.')
            format = kwargs.get('figformat')
            width = kwargs.get('figwidth')
            height = kwargs.get('figheight')
            dpi = kwargs.get('figdpi')

            format = format.lower()
            if figall or cc:
                plt.figure(figsize=(width, height))
                prody.showCrossCorr(pca)
                plt.savefig(join(outdir, prefix + '_cc.' + format),
                            dpi=dpi,
                            format=format)
                plt.close('all')
            if figall or sf:
                plt.figure(figsize=(width, height))
                prody.showSqFlucts(pca)
                plt.savefig(join(outdir, prefix + '_sf.' + format),
                            dpi=dpi,
                            format=format)
                plt.close('all')
            if figall or sp:
                indices = []
                for item in sp.split():
                    try:
                        if '-' in item:
                            item = item.split('-')
                            if len(item) == 2:
                                indices.append(
                                    list(range(int(item[0]) - 1,
                                               int(item[1]))))
                        elif ',' in item:
                            indices.append(
                                [int(i) - 1 for i in item.split(',')])
                        else:
                            indices.append(int(item) - 1)
                    except:
                        pass
                for index in indices:
                    plt.figure(figsize=(width, height))
                    prody.showProjection(ensemble, pca[index])
                    if isinstance(index, Integral):
                        index = [index]
                    index = [str(i + 1) for i in index]
                    plt.savefig(join(
                        outdir,
                        prefix + '_proj_' + '_'.join(index) + '.' + format),
                                dpi=dpi,
                                format=format)
                    plt.close('all')