Topdown

Topdown-related classes. More...

Collaboration diagram for Topdown:

Classes
class	FLASHDeconvAlgorithm
	FLASHDeconv algorithm: ultrafast mass deconvolution algorithm for top down mass spectrometry dataset From MSSpectrum, this class outputs DeconvolvedSpectrum. Deconvolution takes three steps: i) decharging and select candidate masses - speed up via binning ii) collecting isotopes from the candidate masses and deisotope - peak groups are defined here iii) scoring and filter out low scoring masses (i.e., peak groups) More...
class	MassFeatureTrace
	Feature trace in mass dimension for FLASHDeconv This class performs mass tracing on the deconvolved masses by FLASHDeconvAlgorithm In other words, per spectrum deconvolved masses are converted into deconvolved features Currently only works for MS1 spectra. (Top-down DIA is not yet used much). Every time an MS1 spectrum is deconvolved, the relevant information is stored in this class. Tracing is performed at the end of FLASHDeconv run. This class also comes with tsv, TopFD, ProMex format output functions. More...
class	PeakGroup
	Class describing a deconvolved mass. A mass contains multiple (LogMz) peaks of different charges and isotope indices. PeakGroup is the set of such peaks representing a single monoisotopic mass. PeakGroup also contains features that define the quality of it. It is used by Qscore calculation. DeconvolvedSpectrum consists of PeakGroups. More...
class	Qscore
	Qscore : quality score for PeakGroup. This class is being updated. For now, simply it calculate the Qscore using a fixed weight vector. The weight vector has been determined by logistic regression. But afterwards, the training part for the Qscore should be added in here. Or other technique such as deep learning would be used. This class also contains tsv output function. The tsv file contains features of PeakGroups which are used for training. More...
class	Qvalue
	Qvalue : contains functions to calculate Qvalues from deconvolution quality score. More...

Detailed Description

Topdown-related classes.