Grantham (1974) Amino-Acid chemical distance. More...

#include <Bpp/Seq/AlphabetIndex/GranthamAAChemicalDistance.h>

Inheritance diagram for bpp::GranthamAAChemicalDistance:

Collaboration diagram for bpp::GranthamAAChemicalDistance:

Public Member Functions
	GranthamAAChemicalDistance ()

	GranthamAAChemicalDistance (const GranthamAAChemicalDistance &gd)

GranthamAAChemicalDistance &	operator= (const GranthamAAChemicalDistance &gd)

virtual	~GranthamAAChemicalDistance ()

void	setSymmetric (bool yn)

bool	isSymmetric () const

void	setPC1Sign (bool yn)
	The sign of the distance is computed using the coordinate on the first axis of a principal component analysis with the 3 elementary properties (Volume, Polarity, Composition). Otherwise, use the default arbitrary sign. Using this option will lead isSymmetric to return false. More...

Methods from the AlphabetIndex2 interface.
double	getIndex (int state1, int state2) const throw (BadIntException)
	Get the index associated to a pair of states. More...

double	getIndex (const std::string &state1, const std::string &state2) const throw (BadCharException)
	Get the index associated to a pair of states. More...

const Alphabet *	getAlphabet () const
	Get the alphabet associated to this index. More...

GranthamAAChemicalDistance *	clone () const

Matrix< double > *	getIndexMatrix () const

Static Public Attributes
static short int	SIGN_ARBITRARY = 1

static short int	SIGN_PC1 = 2

static short int	SIGN_NONE = 0

Private Attributes
LinearMatrix< double >	distanceMatrix_

LinearMatrix< double >	signMatrix_

const ProteicAlphabet *	alpha_

short int	sign_

Detailed Description

Grantham (1974) Amino-Acid chemical distance.

Two kinds of matrix can be built:

a symmetric one, with $I_{i,j} = I_{i,j}$ ,
or a non-symmetric one, with $I_{i,j} = -I_{i,j}$ . In the second case, which one of the two entries between $I_{i,j}$ and $I_{i,j}$ is positive is arbitrary by default. It is also possible to use the coordinate on a principal component analysis between the elementary propoerties of the distance instead (setPC1Sign(true)). The following R code was use in order to get those signs:
library(seqinr)
data(aaindex)
data<-data.frame(composition=aaindex[["GRAR740101"]]$I,
polarity=aaindex[["GRAR740102"]]$I,
volume=aaindex[["GRAR740103"]]$I)
library(ade4)
pca<-dudi.pca(data)
plot(pca$li[, 1:2], type="n")
text(pca$li[, 1:2], rownames(data))
s.corcircle(pca$co)
layout(matrix(1:3,nrow=1))
a1<-pca$li[,1]; names(a1)<-rownames(data); dotchart(sort(a1))
a2<-pca$li[,2]; names(a2)<-rownames(data); dotchart(sort(a2))
a3<-pca$li[,3]; names(a3)<-rownames(data); dotchart(sort(a3))
x<-pca$li[,1] #Contains the coordinates on the first axis.
m<-matrix(nrow=20, ncol=20)
for(i in 1:length(x))
for(j in 1:length(x))
m[i,j]<-sign(x[j] - x[i])

Reference: Grantham, R. Amino acid difference formula to help explain protein evolution Science 185, 862-864 (1974)

Data from AAIndex2 database, Accession Number GRAR740104.

Definition at line 96 of file GranthamAAChemicalDistance.h.

Constructor & Destructor Documentation

◆ GranthamAAChemicalDistance() [1/2]

GranthamAAChemicalDistance::GranthamAAChemicalDistance ( )

Definition at line 53 of file GranthamAAChemicalDistance.cpp.

Referenced by clone().

◆ GranthamAAChemicalDistance() [2/2]

bpp::GranthamAAChemicalDistance::GranthamAAChemicalDistance ( const GranthamAAChemicalDistance & gd )

inline

Definition at line 108 of file GranthamAAChemicalDistance.h.

◆ ~GranthamAAChemicalDistance()

GranthamAAChemicalDistance::~GranthamAAChemicalDistance ( )

virtual

Definition at line 62 of file GranthamAAChemicalDistance.cpp.

Member Function Documentation

◆ clone()

GranthamAAChemicalDistance* bpp::GranthamAAChemicalDistance::clone ( ) const

inlinevirtual

Implements bpp::AlphabetIndex2.

Definition at line 135 of file GranthamAAChemicalDistance.h.

References GranthamAAChemicalDistance().

◆ getAlphabet()

const Alphabet* bpp::GranthamAAChemicalDistance::getAlphabet ( ) const

inlinevirtual

Get the alphabet associated to this index.

Returns: Alphabet The alphabet associated to this index.

Implements bpp::AlphabetIndex2.

Definition at line 134 of file GranthamAAChemicalDistance.h.

References alpha_.

◆ getIndex() [1/2]

double GranthamAAChemicalDistance::getIndex	(	int	state1,
		int	state2
	)		const
throw	(	BadIntException
	)

virtual

Get the index associated to a pair of states.

Parameters

state1	First state to consider, as a int value.
state2	Second state to consider, as a int value.

Returns: The index associated to the pair of states.

Implements bpp::AlphabetIndex2.

Definition at line 64 of file GranthamAAChemicalDistance.cpp.

◆ getIndex() [2/2]

double GranthamAAChemicalDistance::getIndex	(	const std::string &	state1,
		const std::string &	state2
	)		const
throw	(	BadCharException
	)

virtual

Get the index associated to a pair of states.

Parameters

state1	First state to consider, as a string value.
state2	Second state to consider, as a string value.

Returns: The index associated to the pair of states.

Implements bpp::AlphabetIndex2.

Definition at line 77 of file GranthamAAChemicalDistance.cpp.

◆ getIndexMatrix()

Matrix< double > * GranthamAAChemicalDistance::getIndexMatrix ( ) const

virtual

Returns: A matrix object with all indices.

Implements bpp::AlphabetIndex2.

Definition at line 83 of file GranthamAAChemicalDistance.cpp.

References distanceMatrix_, sign_, SIGN_NONE, SIGN_PC1, and signMatrix_.

◆ isSymmetric()

bool bpp::GranthamAAChemicalDistance::isSymmetric ( ) const

inlinevirtual

Returns: True if the index is symatric (that is, index(i,j) == index(j, i)).

Implements bpp::AlphabetIndex2.

Definition at line 141 of file GranthamAAChemicalDistance.h.

References sign_, and SIGN_NONE.

◆ operator=()

GranthamAAChemicalDistance& bpp::GranthamAAChemicalDistance::operator= ( const GranthamAAChemicalDistance & gd )

inline

Definition at line 115 of file GranthamAAChemicalDistance.h.

References alpha_, distanceMatrix_, sign_, and signMatrix_.

◆ setPC1Sign()

void bpp::GranthamAAChemicalDistance::setPC1Sign ( bool yn )

inline

The sign of the distance is computed using the coordinate on the first axis of a principal component analysis with the 3 elementary properties (Volume, Polarity, Composition). Otherwise, use the default arbitrary sign. Using this option will lead isSymmetric to return false.

Parameters