#include <recombinator_haplotypes.hpp>

Classes
struct	ChildIterator

struct	Iterator

Public Types
enum	kmer_classification : size_t { absent = 0 , heterozygous = 1 , present = 2 , frequent = 3 }
	Kmer classification. More...

typedef std::pair< gbwt::size_type, gbwt::size_type >	sequence_type
	A GBWT sequence as (sequence identifier, offset in a node). More...

typedef std::pair< std::uint32_t, std::uint32_t >	compact_sequence_type
	A more space-efficient representation of `sequence_type`. More...

Public Member Functions
	KmerPresenceMatrix ()
	Default constructor that creates an empty k-mer presence matrix. More...

	KmerPresenceMatrix (size_t num_sequences, size_t num_kmers, sdsl::bit_vector &&matrix)

	KmerPresenceMatrix (const std::vector< sequence_type > &sequences, std::vector< compact_sequence_type > &reordered_sequences, size_t num_kmers, const sdsl::bit_vector &matrix)

	KmerPresenceMatrix (const KmerPresenceMatrix &other)=default

	KmerPresenceMatrix (KmerPresenceMatrix &&other) noexcept=default

KmerPresenceMatrix &	operator= (const KmerPresenceMatrix &other)=default

KmerPresenceMatrix &	operator= (KmerPresenceMatrix &&other) noexcept=default

bool	operator== (const KmerPresenceMatrix &other) const

bool	operator!= (const KmerPresenceMatrix &other) const

size_t	size () const
	Returns the size of the matrix (the number of cells in it). More...

bool	empty () const
	Returns `true` if the matrix is empty. More...

size_t	get_num_sequences () const
	Returns the number of sequences (rows) in the matrix. More...

size_t	get_num_kmers () const
	Returns the number of kmers (columns) in the matrix. More...

size_t	get_num_parents () const

size_t	get_num_present (size_t sequence) const
	Returns the number of present kmers in the given sequence. More...

size_t	get_total_present () const
	Returns the the total number of present kmers in the sequences. More...

void	for_each_kmer (size_t sequence, const std::function< void(size_t, bool)> &callback) const

void	score_sequences (const std::vector< std::pair< kmer_classification, double >> &kmers, const std::function< void(size_t, double)> &sequence_score) const

std::int64_t	diploid_score (size_t first, size_t second, const std::vector< std::pair< kmer_classification, double >> &kmers) const

void	simple_sds_serialize (std::ostream &out) const
	Serializes the object to a stream in the Simple-SDS format. More...

void	simple_sds_load (std::istream &in)
	Loads the object from a stream in the Simple-SDS format. More...

size_t	simple_sds_size () const
	Returns the size of the object in elements. More...

Private Member Functions
size_t	parent (size_t rank) const

size_t	child_rank (size_t rank) const

bool	is_parent (size_t rank) const

Private Attributes
size_t	num_kmers

sdsl::int_vector	parent_rank

sdsl::bit_vector	parents

sdsl::sd_vector	children

Detailed Description

A compressed binary matrix with haplotype sequences as rows and k-mers as columns. Set bits indicate k-mers that are present in the corresponding sequence.

The compression assumes that the matrix consists of clusters of similar sequences. We choose one sequence from each cluster as a parent and store it explicitly. For the other sequences (children), we only store the symmetric difference with their parent sequence. The internal representation reorders the sequences so that the parent of each cluster appears first, followed by its children.

Member Typedef Documentation

◆ compact_sequence_type

typedef std::pair<std::uint32_t, std::uint32_t> vg::KmerPresenceMatrix::compact_sequence_type

A more space-efficient representation of sequence_type.

◆ sequence_type

typedef std::pair<gbwt::size_type, gbwt::size_type> vg::KmerPresenceMatrix::sequence_type

A GBWT sequence as (sequence identifier, offset in a node).

Member Enumeration Documentation

◆ kmer_classification

enum vg::KmerPresenceMatrix::kmer_classification : size_t

Kmer classification.

Enumerator
absent
heterozygous
present
frequent

Constructor & Destructor Documentation

◆ KmerPresenceMatrix() [1/5]

vg::KmerPresenceMatrix::KmerPresenceMatrix ( )

Default constructor that creates an empty k-mer presence matrix.

◆ KmerPresenceMatrix() [2/5]

vg::KmerPresenceMatrix::KmerPresenceMatrix	(	size_t	num_sequences,
		size_t	num_kmers,
		sdsl::bit_vector &&	matrix
	)

Constructor that creates an uncompressed k-mer presence matrix by taking ownership of the given matrix. See below for the semantics of the matrix. Throws std::runtime_error if the dimensions do not match.

◆ KmerPresenceMatrix() [3/5]

vg::KmerPresenceMatrix::KmerPresenceMatrix	(	const std::vector< sequence_type > &	sequences,
		std::vector< compact_sequence_type > &	reordered_sequences,
		size_t	num_kmers,
		const sdsl::bit_vector &	matrix
	)

Constructor from a given uncompressed matrix, where sequence i contains kmer j if and only if matrix[i * num_kmers + j] == 1. Throws std::runtime_error if the dimensions do not match. Reorders the sequences to reordered_sequences to match the internal representation.

◆ KmerPresenceMatrix() [4/5]

vg::KmerPresenceMatrix::KmerPresenceMatrix ( const KmerPresenceMatrix & other )

default

◆ KmerPresenceMatrix() [5/5]

vg::KmerPresenceMatrix::KmerPresenceMatrix ( KmerPresenceMatrix && other )

defaultnoexcept

Member Function Documentation

◆ child_rank()

size_t vg::KmerPresenceMatrix::child_rank ( size_t rank ) const

inlineprivate

◆ diploid_score()

std::int64_t vg::KmerPresenceMatrix::diploid_score	(	size_t	first,
		size_t	second,
		const std::vector< std::pair< kmer_classification, double >> &	kmers
	)		const

Scores a pair of sequences in the subchain. The score is the sum of 1 - abs(expected - observed) over the kmers.

◆ empty()

bool vg::KmerPresenceMatrix::empty ( ) const

inline

Returns true if the matrix is empty.

◆ for_each_kmer()

void vg::KmerPresenceMatrix::for_each_kmer	(	size_t	sequence,
		const std::function< void(size_t, bool)> &	callback
	)		const

Calls callback(kmer_index, is_present) for each kmer in the given sequence, ordered by their indices.

◆ get_num_kmers()

size_t vg::KmerPresenceMatrix::get_num_kmers ( ) const

inline

Returns the number of kmers (columns) in the matrix.

◆ get_num_parents()

size_t vg::KmerPresenceMatrix::get_num_parents ( ) const

inline

Returns the number of clusters, which is also the number of sequences stored explicitly as parents.

◆ get_num_present()

size_t vg::KmerPresenceMatrix::get_num_present ( size_t sequence ) const

Returns the number of present kmers in the given sequence.

◆ get_num_sequences()

size_t vg::KmerPresenceMatrix::get_num_sequences ( ) const

inline

Returns the number of sequences (rows) in the matrix.

◆ get_total_present()

size_t vg::KmerPresenceMatrix::get_total_present ( ) const

Returns the the total number of present kmers in the sequences.

◆ is_parent()

bool vg::KmerPresenceMatrix::is_parent ( size_t rank ) const

inlineprivate

◆ operator!=()

bool vg::KmerPresenceMatrix::operator!= ( const KmerPresenceMatrix & other ) const

inline

◆ operator=() [1/2]

KmerPresenceMatrix& vg::KmerPresenceMatrix::operator= ( const KmerPresenceMatrix & other )

default

◆ operator=() [2/2]

KmerPresenceMatrix& vg::KmerPresenceMatrix::operator= ( KmerPresenceMatrix && other )

defaultnoexcept

◆ operator==()

bool vg::KmerPresenceMatrix::operator== ( const KmerPresenceMatrix & other ) const

◆ parent()

size_t vg::KmerPresenceMatrix::parent ( size_t rank ) const

inlineprivate

◆ score_sequences()

void vg::KmerPresenceMatrix::score_sequences	(	const std::vector< std::pair< kmer_classification, double >> &	kmers,
		const std::function< void(size_t, double)> &	sequence_score
	)		const

Scores every sequence in the subchain. The score of a sequence is the sum of (present : 1 : -1) * kmer.second over kmers. It is reported by calling sequence_score(sequence, score) in an arbitrary order.

NOTE: Because we reuse the score of a parent when scoring its children, the scores are often different from what we would get by scoring each sequence independently. This is due to numerical instability with floating point arithmetic.

◆ simple_sds_load()

void vg::KmerPresenceMatrix::simple_sds_load ( std::istream & in )

Loads the object from a stream in the Simple-SDS format.

◆ simple_sds_serialize()

void vg::KmerPresenceMatrix::simple_sds_serialize ( std::ostream & out ) const

Serializes the object to a stream in the Simple-SDS format.

◆ simple_sds_size()

size_t vg::KmerPresenceMatrix::simple_sds_size ( ) const

Returns the size of the object in elements.

◆ size()

size_t vg::KmerPresenceMatrix::size ( ) const

inline

Returns the size of the matrix (the number of cells in it).

Member Data Documentation

◆ children

sdsl::sd_vector vg::KmerPresenceMatrix::children

private

◆ num_kmers

size_t vg::KmerPresenceMatrix::num_kmers

private

◆ parent_rank

sdsl::int_vector vg::KmerPresenceMatrix::parent_rank

private

◆ parents

sdsl::bit_vector vg::KmerPresenceMatrix::parents

private

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

src/recombinator_haplotypes.hpp
src/recombinator_haplotypes.cpp

Classes

Public Types

Public Member Functions

Private Member Functions

Private Attributes

Detailed Description

Member Typedef Documentation

◆ compact_sequence_type

◆ sequence_type

Member Enumeration Documentation

◆ kmer_classification

Constructor & Destructor Documentation

◆ KmerPresenceMatrix() [1/5]

◆ KmerPresenceMatrix() [2/5]

◆ KmerPresenceMatrix() [3/5]

◆ KmerPresenceMatrix() [4/5]

◆ KmerPresenceMatrix() [5/5]

Member Function Documentation

◆ child_rank()

◆ diploid_score()

◆ empty()

◆ for_each_kmer()

◆ get_num_kmers()

◆ get_num_parents()

◆ get_num_present()

◆ get_num_sequences()

◆ get_total_present()

◆ is_parent()

◆ operator!=()

◆ operator=() [1/2]

◆ operator=() [2/2]

◆ operator==()

◆ parent()

◆ score_sequences()

◆ simple_sds_load()

◆ simple_sds_serialize()

◆ simple_sds_size()

◆ size()

Member Data Documentation

◆ children

◆ num_kmers

◆ parent_rank

◆ parents