This class implements a peak picking algorithm using wavelet techniques. More...

#include <OpenMS/TRANSFORMATIONS/RAW2PEAK/PeakPickerCWT.h>

Inheritance diagram for PeakPickerCWT:

Classes
struct	PeakArea_
	Class for the internal peak representation. More...

Public Types
typedef MSSpectrum::iterator	PeakIterator
	Profile data iterator type. More...

typedef MSSpectrum::const_iterator	ConstPeakIterator
	Const profile data iterator type. More...

Public Types inherited from ProgressLogger
enum	LogType { CMD, GUI, NONE }
	Possible log types. More...

Public Member Functions
	PeakPickerCWT ()
	Constructor. More...

	~PeakPickerCWT () override
	Destructor. More...

void	pick (const MSSpectrum &input, MSSpectrum &output) const
	Applies the peak picking algorithm to a single spectrum. More...

void	pickExperiment (const PeakMap &input, PeakMap &output)
	Picks the peaks in an MSExperiment. More...

double	estimatePeakWidth (const PeakMap &input)
	Estimates average peak width that can then be used for peak picking. More...

Public Member Functions inherited from DefaultParamHandler
	DefaultParamHandler (const String &name)
	Constructor with name that is displayed in error messages. More...

	DefaultParamHandler (const DefaultParamHandler &rhs)
	Copy constructor. More...

virtual	~DefaultParamHandler ()
	Destructor. More...

virtual DefaultParamHandler &	operator= (const DefaultParamHandler &rhs)
	Assignment operator. More...

virtual bool	operator== (const DefaultParamHandler &rhs) const
	Equality operator. More...

void	setParameters (const Param &param)
	Sets the parameters. More...

const Param &	getParameters () const
	Non-mutable access to the parameters. More...

const Param &	getDefaults () const
	Non-mutable access to the default parameters. More...

const String &	getName () const
	Non-mutable access to the name. More...

void	setName (const String &name)
	Mutable access to the name. More...

const std::vector< String > &	getSubsections () const
	Non-mutable access to the registered subsections. More...

Public Member Functions inherited from ProgressLogger
	ProgressLogger ()
	Constructor. More...

	~ProgressLogger ()
	Destructor. More...

	ProgressLogger (const ProgressLogger &other)
	Copy constructor. More...

ProgressLogger &	operator= (const ProgressLogger &other)
	Assignment Operator. More...

void	setLogType (LogType type) const
	Sets the progress log that should be used. The default type is NONE! More...

LogType	getLogType () const
	Returns the type of progress log being used. More...

void	startProgress (SignedSize begin, SignedSize end, const String &label) const
	Initializes the progress display. More...

void	setProgress (SignedSize value) const
	Sets the current progress. More...

void	endProgress () const
	Ends the progress display. More...

void	nextProgress () const
	increment progress by 1 (according to range begin-end) More...

Protected Member Functions
void	updateMembers_ () override
	This method is used to update extra member variables at the end of the setParameters() method. More...

void	getPeakArea_ (const PeakArea_ &area, double &area_left, double &area_right) const
	Computes the peak's left and right area. More...

PeakShape	fitPeakShape_ (const PeakArea_ &area) const
	Returns the best fitting peakshape. More...

double	correlate_ (const PeakShape &peak, const PeakArea_ &area, Int direction=0) const
	Returns the squared Pearson coefficient. More...

bool	getMaxPosition_ (const PeakIterator first, const PeakIterator last, const ContinuousWaveletTransform &wt, PeakArea_ &area, const Int distance_from_scan_border, const double peak_bound_cwt, const double peak_bound_ms2_level_cwt, const Int direction=1) const
	Finds the next maximum position in the wavelet transform wt. More...

bool	getPeakEndPoints_ (PeakIterator first, PeakIterator last, PeakArea_ &area, Int distance_from_scan_border, Int &peak_left_index, Int &peak_right_index, ContinuousWaveletTransformNumIntegration &wt) const
	Determines a peaks's endpoints. More...

void	getPeakCentroid_ (PeakArea_ &area) const
	Estimates a peak's centroid position. More...

double	lorentz_ (const double height, const double lambda, const double pos, const double x) const
	Computes the value of a theoretical Lorentz peak at position x. More...

void	initializeWT_ (ContinuousWaveletTransformNumIntegration &wt, const double peak_bound_in, double &peak_bound_ms_cwt) const
	Computes the threshold for the peak height in the wavelet transform and initializes the wavelet transform. More...

Methods needed for separation of overlapping peaks
bool	deconvolutePeak_ (PeakShape &shape, std::vector< PeakShape > &peak_shapes, double peak_bound_cwt) const
	Separates overlapping peaks. More...

Int	getNumberOfPeaks_ (ConstPeakIterator first, ConstPeakIterator last, std::vector< double > &peak_values, Int direction, double resolution, ContinuousWaveletTransformNumIntegration &wt, double peak_bound_cwt) const
	Determines the number of peaks in the given mass range using the cwt. More...

Int	determineChargeState_ (std::vector< double > &peak_values) const
	Estimate the charge state of the peaks. More...

void	addPeak_ (std::vector< PeakShape > &peaks_DC, PeakArea_ &area, double left_width, double right_width, OptimizePeakDeconvolution::Data &data) const
	Add a peak. More...

Protected Member Functions inherited from DefaultParamHandler
void	defaultsToParam_ ()
	Updates the parameters after the defaults have been set in the constructor. More...

Protected Attributes
float	peak_bound_
	Threshold for the peak height in the MS 1 level. More...

float	peak_bound_ms2_level_
	Threshold for the peak height in the MS 2 level. More...

float	signal_to_noise_
	Signal to noise threshold. More...

float	fwhm_bound_
	The minimal full width at half maximum. More...

UInt	radius_
	The search radius for the determination of a peak's maximum position. More...

float	scale_
	The dilation of the wavelet. More...

float	peak_corr_bound_
	The threshold for correlation. More...

float	noise_level_
	The threshold for the noise level (TODO: Use the information of the signal to noise estimator) More...

bool	optimization_
	Switch for the optimization of peak parameters. More...

bool	deconvolution_
	Switch for the deconvolution of peak parameters. More...

bool	two_d_optimization_
	Switch for the 2D optimization of peak parameters. More...

Protected Attributes inherited from DefaultParamHandler
Param	param_
	Container for current parameters. More...

Param	defaults_
	Container for default parameters. This member should be filled in the constructor of derived classes! More...

std::vector< String >	subsections_
	Container for registered subsections. This member should be filled in the constructor of derived classes! More...

String	error_name_
	Name that is displayed in error messages during the parameter checking. More...

bool	check_defaults_
	If this member is set to false no checking if parameters in done;. More...

bool	warn_empty_defaults_
	If this member is set to false no warning is emitted when defaults are empty;. More...

Protected Attributes inherited from ProgressLogger
LogType	type_

time_t	last_invoke_

ProgressLoggerImpl *	current_logger_

Additional Inherited Members
Static Protected Member Functions inherited from ProgressLogger
static String	logTypeToFactoryName_ (LogType type)
	Return the name of the factory product used for this log type. More...

Static Protected Attributes inherited from ProgressLogger
static int	recursion_depth_

Detailed Description

This class implements a peak picking algorithm using wavelet techniques.

The algorithm is described in detail in Lange et al. (2006) Proc. PSB-06.

This peak picking algorithm uses the continuous wavelet transform of a profile data signal to detect mass peaks. Afterwards a given asymmetric peak function is fitted to the profile data and important peak parameters (e.g. fwhm) are extracted. In an optional step these parameters can be optimized using a non-linear optimization method.

The peak parameters are stored in the meta data arrays of the spectra (see MSSpectrum) in this order:

rValue
area
fwhm
leftWidth
rightWidth
peakShape
SignalToNoise

Note: The peaks must be sorted according to ascending m/z!

Parameters of this class are:

Name	Type	Default	Restrictions	Description
signal_to_noise	float	1.0	min: 0.0	Minimal signal to noise ratio for a peak to be picked.
centroid_percentage	float	0.8	min: 0.0 max: 1.0	Percentage of the maximum height that the raw data points must exceed to be taken into account for the calculation of the centroid. If it is 1 the centroid position corresponds to the position of the highest intensity.
peak_width	float	0.15	min: 0.0	Approximate fwhm of the peaks.
estimate_peak_width	string	false	true, false	Flag if the average peak width shall be estimated. Attention: when this flag is set, the peak_width is ignored.
fwhm_lower_bound_factor	float	0.7	min: 0.0	Factor that calculates the minimal fwhm value from the peak_width. All peaks with width smaller than fwhm_bound_factor * peak_width are discarded.
fwhm_upper_bound_factor	float	20.0	min: 0.0	Factor that calculates the maximal fwhm value from the peak_width. All peaks with width greater than fwhm_upper_bound_factor * peak_width are discarded.
optimization	string	no	no, one_dimensional, two_dimensional	If the peak parameters position, intensity and left/right widthshall be optimized set optimization to one_dimensional or two_dimensional.
thresholds:peak_bound	float	10.0	min: 0.0	Minimal peak intensity.
thresholds:peak_bound_ms2_level	float	10.0	min: 0.0	Minimal peak intensity for MS/MS peaks.
thresholds:correlation	float	0.5	min: 0.0 max: 1.0	minimal correlation of a peak and the raw signal. If a peak has a lower correlation it is skipped.
thresholds:noise_level	float	0.1	min: 0.0	noise level for the search of the peak endpoints.
thresholds:search_radius	int	3	min: 0	search radius for the search of the maximum in the signal after a maximum in the cwt was found
wavelet_transform:spacing	float	0.001	min: 0.0	Spacing of the CWT. Note that the accuracy of the picked peak's centroid position depends in the Raw data spacing, i.e., 50% of raw peak distance at most.
optimization:iterations	int	400	min: 1	maximal number of iterations for the fitting step
optimization:penalties:position	float	0.0	min: 0.0	penalty term for the fitting of the position:If it differs too much from the initial one it can be penalized
optimization:penalties:left_width	float	1.0	min: 0.0	penalty term for the fitting of the left width:If the left width differs too much from the initial one during the fitting it can be penalized.
optimization:penalties:right_width	float	1.0	min: 0.0	penalty term for the fitting of the right width:If the right width differs too much from the initial one during the fitting it can be penalized.
optimization:penalties:height	float	1.0	min: 0.0	penalty term for the fitting of the intensity (only used in 2D Optimization):If it gets negative during the fitting it can be penalized.
optimization:2d:tolerance_mz	float	2.2	min: 0.0	mz tolerance for cluster construction
optimization:2d:max_peak_distance	float	1.2	min: 0.0	maximal peak distance in mz in a cluster
deconvolution:deconvolution	string	false	true, false	If you want heavily overlapping peaks to be separated set this value to "true"
deconvolution:asym_threshold	float	0.3	min: 0.0	If the symmetry of a peak is smaller than asym_thresholds it is assumed that it consists of more than one peak and the deconvolution procedure is started.
deconvolution:left_width	float	2.0	min: 0.0	1/left_width is the initial value for the left width of the peaks found in the deconvolution step.
deconvolution:right_width	float	2.0	min: 0.0	1/right_width is the initial value for the right width of the peaks found in the deconvolution step.
deconvolution:scaling	float	0.12	min: 0.0	Initial scaling of the cwt used in the separation of heavily overlapping peaks. The initial value is used for charge 1, for higher charges it is adapted to scaling/charge.
deconvolution:fitting:fwhm_threshold	float	0.7	min: 0.0	If the FWHM of a peak is higher than 'fwhm_thresholds' it is assumed that it consists of more than one peak and the deconvolution procedure is started.
deconvolution:fitting:eps_abs	float	9.999999747378752e-06	min: 0.0	if the absolute error gets smaller than this value the fitting is stopped.
deconvolution:fitting:eps_rel	float	9.999999747378752e-06	min: 0.0	if the relative error gets smaller than this value the fitting is stopped.
deconvolution:fitting:max_iteration	int	10	min: 1	maximal number of iterations for the fitting step
deconvolution:fitting:penalties:position	float	0.0	min: 0.0	penalty term for the fitting of the peak position:If the position changes more than 0.5Da during the fitting it can be penalized as well as discrepancies of the peptide mass rule.
deconvolution:fitting:penalties:height	float	1.0	min: 0.0	penalty term for the fitting of the intensity:If it gets negative during the fitting it can be penalized.
deconvolution:fitting:penalties:left_width	float	0.0	min: 0.0	penalty term for the fitting of the left width:If the left width gets too broad or negative during the fitting it can be penalized.
deconvolution:fitting:penalties:right_width	float	0.0	min: 0.0	penalty term for the fitting of the right width:If the right width gets too broad or negative during the fitting it can be penalized.
SignalToNoiseEstimationParameter:max_intensity	int	-1	min: -1	maximal intensity considered for histogram construction. By default, it will be calculated automatically (see auto_mode). Only provide this parameter if you know what you are doing (and change 'auto_mode' to '-1')! All intensities EQUAL/ABOVE 'max_intensity' will not be added to the histogram. If you choose 'max_intensity' too small, the noise estimate might be too small as well. If chosen too big, the bins become quite large (which you could counter by increasing 'bin_count', which increases runtime).
SignalToNoiseEstimationParameter:auto_max_stdev_factor	float	3.0	min: 0.0 max: 999.0	parameter for 'max_intensity' estimation (if 'auto_mode' == 0): mean + 'auto_max_stdev_factor' * stdev
SignalToNoiseEstimationParameter:auto_max_percentile	int	95	min: 0 max: 100	parameter for 'max_intensity' estimation (if 'auto_mode' == 1): auto_max_percentile th percentile
SignalToNoiseEstimationParameter:auto_mode	int	0	min: -1 max: 1	method to use to determine maximal intensity: -1 --> use 'max_intensity'; 0 --> 'auto_max_stdev_factor' method (default); 1 --> 'auto_max_percentile' method
SignalToNoiseEstimationParameter:win_len	float	200.0	min: 1.0	window length in Thomson
SignalToNoiseEstimationParameter:bin_count	int	30	min: 3	number of bins for intensity values
SignalToNoiseEstimationParameter:stdev_mp	float	3.0	min: 0.01 max: 999.0	multiplier for stdev
SignalToNoiseEstimationParameter:min_required_elements	int	10	min: 1	minimum number of elements required in a window (otherwise it is considered sparse)
SignalToNoiseEstimationParameter:noise_for_empty_window	float	1.0e20		noise value used for sparse windows

Note:

If a section name is documented, the documentation is displayed as tooltip.
Advanced parameter names are italic.

Class Documentation

◆ OpenMS::PeakPickerCWT::PeakArea_

struct OpenMS::PeakPickerCWT::PeakArea_

Class for the internal peak representation.

A regular Data-Object which contains some additional useful information for analyzing peaks and their properties The left and right iterators delimit a range in the profile data which represents a profile peak. They define the profile peak endpoints. max points to the profile data point in [left, right] with the highest intensity, the maximum of the profile peak.

Class Members
typedef iterator	PeakIterator

Class Members
DPosition< 1 >	centroid_position	The estimated centroid position in m/z.
PeakIterator	left	iterator to the leftmost valid point
PeakIterator	max	iterator to the maximum position
PeakIterator	right	iterator to the rightmost valid point (inclusive)

Member Typedef Documentation

◆ ConstPeakIterator

typedef MSSpectrum::const_iterator ConstPeakIterator

Const profile data iterator type.

◆ PeakIterator

typedef MSSpectrum::iterator PeakIterator

Profile data iterator type.

Constructor & Destructor Documentation

◆ PeakPickerCWT()

PeakPickerCWT ( )

Constructor.

◆ ~PeakPickerCWT()

~PeakPickerCWT ( )

override

Destructor.

Member Function Documentation

◆ addPeak_()

void addPeak_	(	std::vector< PeakShape > &	peaks_DC,
		PeakArea_ &	area,
		double	left_width,
		double	right_width,
		OptimizePeakDeconvolution::Data &	data
	)		const

protected

Add a peak.

◆ correlate_()

double correlate_	(	const PeakShape &	peak,
		const PeakArea_ &	area,
		Int	direction = `0`
	)		const

protected

Returns the squared Pearson coefficient.

Computes the correlation of the peak and the original data given by the peak endpoints area.left and area.right. If the value is near 1, the fitted peakshape and the profile data are expected to be very similar.

◆ deconvolutePeak_()

bool deconvolutePeak_	(	PeakShape &	shape,
		std::vector< PeakShape > &	peak_shapes,
		double	peak_bound_cwt
	)		const

protected

Separates overlapping peaks.

It determines the number of peaks lying underneath the initial peak using the cwt with different scales. Then a nonlinear optimization procedure is applied to optimize the peak parameters.

◆ determineChargeState_()

Int determineChargeState_ ( std::vector< double > & peak_values ) const

protected

Estimate the charge state of the peaks.

◆ estimatePeakWidth()

double estimatePeakWidth ( const PeakMap & input )

Estimates average peak width that can then be used for peak picking.

The spectra with the highest TICs are used to estimate an average peak width that can be used as the peak_width parameter for picking the complete data set. Typically, the number of peaks increases with decreasing peak width until a plateau is reached. The beginning of this plateau is our estimate for the peak width. This estimate is averaged over several spectra.

◆ fitPeakShape_()

PeakShape fitPeakShape_ ( const PeakArea_ & area ) const

protected

Returns the best fitting peakshape.

◆ getMaxPosition_()

bool getMaxPosition_	(	const PeakIterator	first,
		const PeakIterator	last,
		const ContinuousWaveletTransform &	wt,
		PeakArea_ &	area,
		const Int	distance_from_scan_border,
		const double	peak_bound_cwt,
		const double	peak_bound_ms2_level_cwt,
		const Int	direction = `1`
	)		const

protected

Finds the next maximum position in the wavelet transform wt.

If the maximum is greater than peak_bound_cwt we search for the corresponding maximum in the profile data interval [first,last) given a predefined search radius radius. Only peaks with intensities greater than peak_bound_ are relevant. If no peak is detected the method return false. For direction=1, the method runs from first to last given direction=-1 it runs the other way around.

◆ getNumberOfPeaks_()

Int getNumberOfPeaks_	(	ConstPeakIterator	first,
		ConstPeakIterator	last,
		std::vector< double > &	peak_values,
		Int	direction,
		double	resolution,
		ContinuousWaveletTransformNumIntegration &	wt,
		double	peak_bound_cwt
	)		const

protected

Determines the number of peaks in the given mass range using the cwt.

◆ getPeakArea_()

void getPeakArea_	(	const PeakArea_ &	area,
		double &	area_left,
		double &	area_right
	)		const

protected

Computes the peak's left and right area.

◆ getPeakCentroid_()

void getPeakCentroid_ ( PeakArea_ & area ) const

protected

Estimates a peak's centroid position.

Computes the centroid position of the peak using all profile data points which are greater than 'centroid_percentage' (user-param) of the most intensive profile data point.

◆ getPeakEndPoints_()

bool getPeakEndPoints_	(	PeakIterator	first,
		PeakIterator	last,
		PeakArea_ &	area,
		Int	distance_from_scan_border,
		Int &	peak_left_index,
		Int &	peak_right_index,
		ContinuousWaveletTransformNumIntegration &	wt
	)		const

protected

Determines a peaks's endpoints.

The algorithm does the following:

let x_m be the position of the maximum in the data and let (x_l, x_r) be the left and right neighbours
(1) starting from x_l', walk left until one of the following happens
the new point is lower than the original bound => we found our left endpoint
the new point is larger than the last, but the point left from the new point is smaller. In that case, we either ran into another peak, or we encounter some noise. Therefore we now look in the cwt at the position corresponding to this value. If the cwt here is monotonous, we consider the point as noise and continue further to the left. Otherwise, we probably found the beginning of a new peak and therefore stop here.

(2) analogous procedure to the right of x_r

◆ initializeWT_()

void initializeWT_	(	ContinuousWaveletTransformNumIntegration &	wt,
		const double	peak_bound_in,
		double &	peak_bound_ms_cwt
	)		const

protected

Computes the threshold for the peak height in the wavelet transform and initializes the wavelet transform.

Given the threshold for the peak height a corresponding value peak_bound_cwt can be computed for the continuous wavelet transform. Therefore we compute a theoretical Lorentzian peakshape with height=peak_bound_ and a width which is similar to the width of the wavelet. Taking the maximum in the wavelet transform of the Lorentzian peak we have a peak bound in the wavelet transform.

◆ lorentz_()

double lorentz_	(	const double	height,
		const double	lambda,
		const double	pos,
		const double	x
	)		const

inlineprotected

Computes the value of a theoretical Lorentz peak at position x.

◆ pick()

void pick	(	const MSSpectrum &	input,
		MSSpectrum &	output
	)		const

Applies the peak picking algorithm to a single spectrum.

Picks the peaks in the input spectrum and writes the resulting peaks to the output container.

◆ pickExperiment()

void pickExperiment	(	const PeakMap &	input,
		PeakMap &	output
	)

Picks the peaks in an MSExperiment.

Picks the peaks successive in every scan in the spectrum range. The detected peaks are stored in the output MSExperiment.

Exceptions

Exception::UnableToFit() if peak width cannot be determined (if estimation is set to auto)

◆ updateMembers_()

void updateMembers_ ( )

overrideprotectedvirtual

This method is used to update extra member variables at the end of the setParameters() method.

Also call it at the end of the derived classes' copy constructor and assignment operator.

The default implementation is empty.

Reimplemented from DefaultParamHandler.

Member Data Documentation

◆ deconvolution_

bool deconvolution_

protected

Switch for the deconvolution of peak parameters.

◆ fwhm_bound_

float fwhm_bound_

protected

The minimal full width at half maximum.

◆ noise_level_

float noise_level_

protected

The threshold for the noise level (TODO: Use the information of the signal to noise estimator)

◆ optimization_

bool optimization_

protected

Switch for the optimization of peak parameters.

◆ peak_bound_

float peak_bound_

protected

Threshold for the peak height in the MS 1 level.

◆ peak_bound_ms2_level_

float peak_bound_ms2_level_

protected

Threshold for the peak height in the MS 2 level.

◆ peak_corr_bound_

float peak_corr_bound_

protected

The threshold for correlation.

◆ radius_

UInt radius_

protected

The search radius for the determination of a peak's maximum position.

◆ scale_

float scale_

protected

The dilation of the wavelet.

◆ signal_to_noise_

float signal_to_noise_

protected

Signal to noise threshold.

◆ two_d_optimization_

bool two_d_optimization_

protected

Switch for the 2D optimization of peak parameters.

Classes

Public Types

Public Member Functions

Protected Member Functions

Protected Attributes

Additional Inherited Members

Detailed Description

Class Documentation

◆ OpenMS::PeakPickerCWT::PeakArea_

Member Typedef Documentation

◆ ConstPeakIterator

◆ PeakIterator

Constructor & Destructor Documentation

◆ PeakPickerCWT()

◆ ~PeakPickerCWT()

Member Function Documentation

◆ addPeak_()

◆ correlate_()

◆ deconvolutePeak_()

◆ determineChargeState_()

◆ estimatePeakWidth()

◆ fitPeakShape_()

◆ getMaxPosition_()

◆ getNumberOfPeaks_()

◆ getPeakArea_()

◆ getPeakCentroid_()

◆ getPeakEndPoints_()

◆ initializeWT_()

◆ lorentz_()

◆ pick()

◆ pickExperiment()

◆ updateMembers_()

Member Data Documentation

◆ deconvolution_

◆ fwhm_bound_

◆ noise_level_

◆ optimization_

◆ peak_bound_

◆ peak_bound_ms2_level_

◆ peak_corr_bound_

◆ radius_

◆ scale_

◆ signal_to_noise_

◆ two_d_optimization_