Similarity score of SpectraST. More...
#include <OpenMS/COMPARISON/SPECTRA/SpectraSTSimilarityScore.h>
Public Member Functions | |
| SpectraSTSimilarityScore & | operator= (const SpectraSTSimilarityScore &source) |
| assignment operator | |
| DoubleReal | operator() (const PeakSpectrum &spec1, const PeakSpectrum &spec2) const |
| : calculates the dot product of the two spectra | |
| DoubleReal | operator() (const BinnedSpectrum &bin1, const BinnedSpectrum &bin2) const |
| : calculates the dot product of the two spectra | |
| DoubleReal | operator() (const PeakSpectrum &spec) const |
| : calculates the dot product of itself | |
| bool | preprocess (PeakSpectrum &spec, Real remove_peak_intensity_threshold=2.01, UInt cut_peaks_below=1000, Size min_peak_number=5, Size max_peak_number=150) |
| Preprocesses the spectrum. | |
| BinnedSpectrum | transform (const PeakSpectrum &spec) |
| spectrum is transformed into a binned spectrum with bin size 1 and spread 1 and the intensities are normalized. | |
| DoubleReal | dot_bias (const BinnedSpectrum &bin1, const BinnedSpectrum &bin2, DoubleReal dot_product=-1) const |
| Calculates how much of the dot prudct is dominated by a few peaks. | |
| DoubleReal | delta_D (DoubleReal top_hit, DoubleReal runner_up) |
| calculates the normalized distance between top_hit and runner_up. | |
| DoubleReal | compute_F (DoubleReal dot_product, DoubleReal delta_D, DoubleReal dot_bias) |
| : computes the overall all score | |
| SpectraSTSimilarityScore () | |
| default constructor | |
| SpectraSTSimilarityScore (const SpectraSTSimilarityScore &source) | |
| copy constructor | |
| virtual | ~SpectraSTSimilarityScore () |
| destructor | |
Static Public Member Functions | |
| static PeakSpectrumCompareFunctor * | create () |
| static const String | getProductName () |
| Reimplemented from PeakSpectrumCompareFunctor. | |
Detailed Description
Similarity score of SpectraST.
Unlike the other similarity scores this score is used for matching a spectrum against a whole library, although the dot poduct seems to be an effective method for scoring on its own. For calculating the Spectrast score, first preprocess the spectra if not already done. Transform them and calculate the dot product and the dot bias. Afterwards get the best two hits and calculate delta_D. Now for every spectrum from the library you can calculate the final score.
The details of the score can be found in: H. Lam et al., Development and validation of a spectral library searching method for peptide identification from MS/MS, Proteomics, 7 , 655-667, 2007
Constructor & Destructor Documentation
default constructor
| SpectraSTSimilarityScore | ( | const SpectraSTSimilarityScore & | source | ) |
copy constructor
| virtual ~SpectraSTSimilarityScore | ( | ) | [virtual] |
destructor
Member Function Documentation
| DoubleReal compute_F | ( | DoubleReal | dot_product, | |
| DoubleReal | delta_D, | |||
| DoubleReal | dot_bias | |||
| ) |
: computes the overall all score
- Parameters:
-
dot_product of a match delta_D should be calculated after all dot products for a unidentified spectrum are computed dot_bias
- Returns:
- the SpectraST similarity score
| static PeakSpectrumCompareFunctor* create | ( | ) | [inline, static] |
| DoubleReal delta_D | ( | DoubleReal | top_hit, | |
| DoubleReal | runner_up | |||
| ) |
calculates the normalized distance between top_hit and runner_up.
- Parameters:
-
top_hit is the best score for a given match. runner_up a match with a worse score than top_hit. e.g. the second best score.
- Returns:
- normalized distance
- Exceptions:
-
DividedByZero exception if top_hit is 0.
- Note:
- Range of the dot products is between 0 and 1.
| DoubleReal dot_bias | ( | const BinnedSpectrum & | bin1, | |
| const BinnedSpectrum & | bin2, | |||
| DoubleReal | dot_product = -1 | |||
| ) | const |
Calculates how much of the dot prudct is dominated by a few peaks.
- Parameters:
-
dot_product if -1 this value will be calculated as well. bin1 first spectrum in binned representation bin2 second spectrum in binned representation
| static const String getProductName | ( | ) | [inline, static] |
Reimplemented from PeakSpectrumCompareFunctor.
Reimplemented from PeakSpectrumCompareFunctor.
| DoubleReal operator() | ( | const BinnedSpectrum & | bin1, | |
| const BinnedSpectrum & | bin2 | |||
| ) | const |
: calculates the dot product of the two spectra
| DoubleReal operator() | ( | const PeakSpectrum & | spec | ) | const [virtual] |
: calculates the dot product of itself
Implements PeakSpectrumCompareFunctor.
| DoubleReal operator() | ( | const PeakSpectrum & | spec1, | |
| const PeakSpectrum & | spec2 | |||
| ) | const [virtual] |
: calculates the dot product of the two spectra
Implements PeakSpectrumCompareFunctor.
| SpectraSTSimilarityScore& operator= | ( | const SpectraSTSimilarityScore & | source | ) |
assignment operator
| bool preprocess | ( | PeakSpectrum & | spec, | |
| Real | remove_peak_intensity_threshold = 2.01, |
|||
| UInt | cut_peaks_below = 1000, |
|||
| Size | min_peak_number = 5, |
|||
| Size | max_peak_number = 150 | |||
| ) |
Preprocesses the spectrum.
The preprocessing removes peak below a intensity threshold, reject spectra that does not have enough peaks, and cuts peaks exceeding the max_peak_number most intense peaks.
- Returns:
- true if spectrum passes filtering
| BinnedSpectrum transform | ( | const PeakSpectrum & | spec | ) |
spectrum is transformed into a binned spectrum with bin size 1 and spread 1 and the intensities are normalized.
| OpenMS / TOPP release 1.10.0 | Documentation generated on Thu Mar 7 2013 09:42:51 using doxygen 1.7.1 |