Re-quantifies linked features at one elution width per precursor. More...

#include <OpenMS/ANALYSIS/QUANTITATION/ConsensusElutionWidth.h>

Classes
struct	Stats

Static Public Member Functions
static Stats	apply (ConsensusMap &consensus, const std::vector< FeatureMap > &feature_maps, const std::vector< TransformationDescription > &transformations)
	Re-quantify the feature handles of `consensus`.

static double	localSlope (const TransformationDescription &trafo, double rt)
	Slope of `trafo` at `rt` by a central difference over +- 1 s (1 for an identity or empty transformation)

Static Public Attributes
static constexpr const char *	FIT_META_VALUES [] = {"model_height", "model_Gauss_sigma", "model_center", "model_status"}
	feature meta values of the elution model fit that apply() reads (written by ElutionModelFitter)

Detailed Description

Re-quantifies linked features at one elution width per precursor.

FeatureFinderIdentification fits an elution model to each feature and reports its area, height x sigma x sqrt(2 pi) for the Gaussian model. The width is fitted per run, so its error moves the area of a precursor from run to run although the peak height is right. This class replaces each run's width by one consensus width per consensus feature:

w_c = median over the identified features of the consensus feature of sigma_r * g_r'(t_r)

where t_r is the feature's apex in run r (meta value "model_center", run RT) and g_r' the slope of the RT transformation of run r (run RT to consensus RT) at t_r, which carries the width onto the consensus RT scale. Every feature of the consensus feature with a valid fit then reports

height_r x (w_c / g_r'(t_r)) x sqrt(2 pi).

A feature is identified if it carries a peptide identification of its own run. Features without a valid Gaussian fit (meta values "model_height", "model_Gauss_sigma", "model_status" starting with '0') keep their intensity, and so do all features of a consensus feature without identified fitted features.

Ported from bigbio/andes #150 (finalize: one consensus width per precursor).

Class Documentation

◆ OpenMS::ConsensusElutionWidth::Stats

struct OpenMS::ConsensusElutionWidth::Stats

Class Members
Size	consensus_features = 0	consensus features with a consensus width
Size	handles_kept = 0	feature handles that kept their intensity
Size	handles_updated = 0	feature handles re-quantified

Member Function Documentation

◆ apply()

static Stats apply	(	ConsensusMap &	consensus,
		const std::vector< FeatureMap > &	feature_maps,
		const std::vector< TransformationDescription > &	transformations
	)

static

Re-quantify the feature handles of consensus.

Parameters

consensus	The linked map; its handle intensities are replaced
feature_maps	The feature maps that were linked, indexed by map index (features are looked up by unique id)
transformations	RT transformations (run RT to consensus RT) per map; empty or shorter: slope 1 for the rest

◆ localSlope()

static double localSlope	(	const TransformationDescription &	trafo,
		double	rt
	)

static

Slope of trafo at rt by a central difference over +- 1 s (1 for an identity or empty transformation)

Member Data Documentation

◆ FIT_META_VALUES

constexpr const char* FIT_META_VALUES[] = {"model_height", "model_Gauss_sigma", "model_center", "model_status"}

staticconstexpr

feature meta values of the elution model fit that apply() reads (written by ElutionModelFitter)

Classes

Static Public Member Functions

Static Public Attributes