OpenMS
2.5.0
|
Predict peak heights of peptides based on Local Linear Map model. More...
#include <OpenMS/ANALYSIS/PIP/PeakIntensityPredictor.h>
Public Member Functions | |
PeakIntensityPredictor () | |
Constructors and Destructors. More... | |
virtual | ~PeakIntensityPredictor () |
destructor More... | |
double | predict (const AASequence &sequence) |
Returns predicted peak heights (intensities) of a single peptide. More... | |
double | predict (const AASequence &sequence, std::vector< double > &add_info) |
Returns predicted peak heights (intensities) of a single peptide. More... | |
std::vector< double > | predict (const std::vector< AASequence > &sequences) |
Returns predicted peak heights (intensities) of several peptides. More... | |
std::vector< double > | predict (const std::vector< AASequence > &sequences, std::vector< std::vector< double > > &add_info) |
Returns predicted peak heights (intensities) of several peptides. More... | |
Private Member Functions | |
double | map_ (const std::vector< double > &data) |
calculate and return predicted value based on given LocalLinearMap model for corresponding aaindex variables More... | |
Size | findWinner_ (const std::vector< double > &data) |
find winning prototype More... | |
std::vector< double > | calculateAddInfo_ (const std::vector< double > &data) |
calculate assignments of peptides to cluster and the corresponding error More... | |
std::vector< double > | getPropertyVector_ (const AASequence &sequence) |
Calculates an array of properties for an amino acid sequence. More... | |
PeakIntensityPredictor (const PeakIntensityPredictor &llmModel) | |
copy constructor not implemented => private More... | |
PeakIntensityPredictor & | operator= (const PeakIntensityPredictor &peakIntensityPredictor) |
assignment operator not implemented => private More... | |
Private Attributes | |
LocalLinearMap | llm_ |
Local Linear Map model. More... | |
Predict peak heights of peptides based on Local Linear Map model.
This class can be used for predictions of peptide peak heights (referred to as intensities) from a peptide sequence by a Local Linear Map (LLM) model. A general introduction to the Peak Intensity Predictor (PIP) can be found in the PIP Tutorial.
The predictor performs only on the peptides sequences as an AASequence representation. Every sequence is transformed to an 18 dimensional data vector representing certain chemical characteristics and is loaded into the trained LocalLinearMap model to find the predicted peptides peak intensity.
Every predictor object calls the appropriate LocalLinearMap model, transforms the given sequences and creates a vector space in which the LocalLinearMap performs.
Constructors and Destructors.
default constructor
|
virtual |
destructor
|
private |
copy constructor not implemented => private
calculate assignments of peptides to cluster and the corresponding error
|
private |
Calculates an array of properties for an amino acid sequence.
The array contains the following properties:
InvalidValue | is thrown if an undefined one-letter-code is used |
calculate and return predicted value based on given LocalLinearMap model for corresponding aaindex variables
|
private |
assignment operator not implemented => private
double predict | ( | const AASequence & | sequence | ) |
Returns predicted peak heights (intensities) of a single peptide.
double predict | ( | const AASequence & | sequence, |
std::vector< double > & | add_info | ||
) |
Returns predicted peak heights (intensities) of a single peptide.
Some additional information is returned in add_info
:
std::vector<double> predict | ( | const std::vector< AASequence > & | sequences | ) |
Returns predicted peak heights (intensities) of several peptides.
std::vector<double> predict | ( | const std::vector< AASequence > & | sequences, |
std::vector< std::vector< double > > & | add_info | ||
) |
Returns predicted peak heights (intensities) of several peptides.
Some additional information for each peptide is returned in add_info
. For each peptide a row with the following components is returned:
|
private |
Local Linear Map model.