Aligning two sequences. More...

#include <yat/utility/Aligner.h>

Classes
class	Cigar
	Compact Idiosyncratic Gapped Alignment Report. More...

Public Types
enum	direction { none, right, down, diagonal }

Public Member Functions
	Aligner (double gap=0, double open_gap=0)
	Constructor. More...

	Aligner (double vertical_gap, double open_vertical_gap, double horizon_gap, double open_horizon_gap)
	constructor More...

double	needleman_wunsch (const MatrixBase &d)

double	smith_waterman (const MatrixBase &d)

void	operator() (const MatrixBase &dot, MatrixMutable &score)
	calculate score matrix based on the dot matrix More...

const direction &	alignment (size_t i, size_t j) const

const Cigar	cigar (size_t i, size_t j) const

Detailed Description

Aligning two sequences.

General class aligning two sequences described in a dot-matrix, D, in which $ D(i,j) $ describes how similar element i in the first sequence is with element j in the second sequence. A path through this matrix corresponds to an alignment of the two sequences, in which a diagonal step correspoinds to a match between corresponding sequence elements, and a vertical or horizontal step correspond correspond to inserting a gap into one of the sequences.

Since: New in yat 0.9

Member Enumeration Documentation

◆ direction

enum theplu::yat::utility::Aligner::direction

Enum used to describe alignment.

See also: alignment

Constructor & Destructor Documentation

◆ Aligner() [1/2]

theplu::yat::utility::Aligner::Aligner	(	double	gap = `0`,
		double	open_gap = `0`
	)

explicit

Constructor.

Same as Aligner(gap, open_gap, gap, open_gap)

◆ Aligner() [2/2]

theplu::yat::utility::Aligner::Aligner	(	double	vertical_gap,
		double	open_vertical_gap,
		double	horizon_gap,
		double	open_horizon_gap
	)

constructor

Parameters

vertical_gap	Penalty for extending a vertical gap. A vertical gap means alignment consumes an element in second element and not, in the first element, i.e., an element has been inserted to the second element or equivalently an element has been deleted in first sequence.
open_vertical_gap	Penalty for open a vertical gap. Total cost for a vertical gap is open_vertical_gap + N * vertical_gap.
horizon_gap	Penalty for extending a horizontal gap
open_horizon_gap	Penalty for open a horizontal gap. Total penalty for a horizontal gap is open_horizon_gap + N horizon_gap.

Member Function Documentation

◆ alignment()

const direction& theplu::yat::utility::Aligner::alignment	(	size_t	i,
		size_t	j
	)		const

If, for example, alignment(i,j) returns Aligner::diagonal, implies that score(i,j) was maximized as score(i-1,j-1) + dot(i-1, j-1), in other words, element i-1 in first sequence was aligned with element j-1 in second sequence.

◆ cigar()

const Cigar theplu::yat::utility::Aligner::cigar	(	size_t	i,
		size_t	j
	)		const

Since: new in yat 0.12

◆ needleman_wunsch()

double theplu::yat::utility::Aligner::needleman_wunsch ( const MatrixBase & d )

Initialize a score matrix with $S_{k,0} = - \textrm{open\_vertical\_gap} - k * \textrm{vertical\_gap}$ and $S_{0,k} = - \textrm{open\_horizon\_gap} - k * \textrm{horizon\_gap}$ and align using operator().

Returns: most lower right element of score matrix

◆ operator()()

void theplu::yat::utility::Aligner::operator()	(	const MatrixBase &	dot,
		MatrixMutable &	score
	)

calculate score matrix based on the dot matrix

The algorithm calculates a maximum score matrix as

$S_{i,j} = \textrm{max} \{ S_{ij}; S_{i,j-1} - F(\textrm{gap}_V); S_{i-1,j-1} + D_{i-1,j-1}; S_{i-1,j} - F(\textrm{gap}_H) \}$

where $F(\textrm{gap})$ is gap if there was already a gap, and gap + open_gap otherwise.

To get the CIGAR to behave as expected, the each row in dot corresponds to an element in reference sequence and each row in dot corresponds to an element in query reference. If that is transposed, insertions and deletions are swapped in CIGAR.

The score matrix must have one column more than the dot matrix and one more row than the dot matrix.

◆ smith_waterman()

double theplu::yat::utility::Aligner::smith_waterman ( const MatrixBase & d )

Initialize a score matrix with all elements equal to zero, align using operator().

Returns: max element in score matrix.

See also: SmithWaterman

The documentation for this class was generated from the following file:

yat/utility/Aligner.h

Classes

Public Types

Public Member Functions