vg::HaplotypePartitioner Class Reference

#include <recombinator.hpp>

Classes
struct	Parameters
	Parameters for partition_haplotypes(). More...
struct	Subchain

Public Types
typedef Haplotypes::Verbosity	Verbosity
	The amount of progress information that should be printed to stderr. More...
typedef Haplotypes::sequence_type	sequence_type
	A GBWT sequence as (sequence identifier, offset in a node). More...
typedef Haplotypes::Subchain::kmer_type	kmer_type
	An encoded kmer. More...
typedef gbwtgraph::MinimizerIndex< gbwtgraph::Key64, gbwtgraph::Position >	minimizer_index_type
	Minimizer index without payloads. More...

Public Member Functions
	HaplotypePartitioner (const gbwtgraph::GBZ &gbz, const gbwt::FastLocate &r_index, const SnarlDistanceIndex &distance_index, const minimizer_index_type &minimizer_index, Verbosity verbosity)
	Creates a new HaplotypePartitioner using the given indexes. More...
Haplotypes	partition_haplotypes (const Parameters &parameters) const

Public Attributes
const gbwtgraph::GBZ &	gbz
const gbwt::FastLocate &	r_index
const SnarlDistanceIndex &	distance_index
const minimizer_index_type &	minimizer_index
Verbosity	verbosity

Static Public Attributes
constexpr static size_t	SUBCHAIN_LENGTH = 10000
	Target length of a subchain. More...
constexpr static size_t	APPROXIMATE_JOBS = 32
	Approximate number of construction jobs to be created. More...

Private Member Functions
size_t	get_distance (handle_t from, handle_t to) const
bool	contains_reversals (handle_t handle) const
std::vector< Subchain >	get_subchains (const gbwtgraph::TopLevelChain &chain, const Parameters &parameters) const
std::vector< sequence_type >	get_sequence_visits (handle_t handle) const
std::vector< sequence_type >	get_sequences (handle_t handle) const
std::vector< sequence_type >	get_sequences (Subchain subchain) const
std::vector< kmer_type >	unique_minimizers (gbwt::size_type sequence_id) const
std::vector< kmer_type >	unique_minimizers (sequence_type sequence, Subchain subchain) const
void	build_subchains (const gbwtgraph::TopLevelChain &chain, Haplotypes::TopLevelChain &output, const Parameters &parameters) const

Detailed Description

A tool for transforming the haplotypes in a GBWT index into a Haplotypes representation. Requires a GBZ graph, an r-index, a distance index, and a minimizer index.

Member Typedef Documentation

kmer_type

typedef Haplotypes::Subchain::kmer_type vg::HaplotypePartitioner::kmer_type

An encoded kmer.

minimizer_index_type

typedef gbwtgraph::MinimizerIndex<gbwtgraph::Key64, gbwtgraph::Position> vg::HaplotypePartitioner::minimizer_index_type

Minimizer index without payloads.

sequence_type

typedef Haplotypes::sequence_type vg::HaplotypePartitioner::sequence_type

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

Verbosity

typedef Haplotypes::Verbosity vg::HaplotypePartitioner::Verbosity

The amount of progress information that should be printed to stderr.

Constructor & Destructor Documentation

HaplotypePartitioner()

vg::HaplotypePartitioner::HaplotypePartitioner	(	const gbwtgraph::GBZ &	gbz,
		const gbwt::FastLocate &	r_index,
		const SnarlDistanceIndex &	distance_index,
		const minimizer_index_type &	minimizer_index,
		Verbosity	verbosity
	)

Creates a new HaplotypePartitioner using the given indexes.

Member Function Documentation

build_subchains()

void vg::HaplotypePartitioner::build_subchains ( const gbwtgraph::TopLevelChain &  chain, Haplotypes::TopLevelChain &  output, const Parameters &  parameters  ) const

private

contains_reversals()

bool vg::HaplotypePartitioner::contains_reversals ( handle_t  handle ) const

private

get_distance()

size_t vg::HaplotypePartitioner::get_distance ( handle_t  from, handle_t  to  ) const

private

get_sequence_visits()

std::vector< HaplotypePartitioner::sequence_type > vg::HaplotypePartitioner::get_sequence_visits ( handle_t  handle ) const

private

get_sequences() [1/2]

std::vector< HaplotypePartitioner::sequence_type > vg::HaplotypePartitioner::get_sequences ( handle_t  handle ) const

private

get_sequences() [2/2]

std::vector< HaplotypePartitioner::sequence_type > vg::HaplotypePartitioner::get_sequences ( Subchain  subchain ) const

private

get_subchains()

std::vector< HaplotypePartitioner::Subchain > vg::HaplotypePartitioner::get_subchains ( const gbwtgraph::TopLevelChain &  chain, const Parameters &  parameters  ) const

private

partition_haplotypes()

Haplotypes vg::HaplotypePartitioner::partition_haplotypes

(

const Parameters &

parameters

)

const

Creates a Haplotypes representation of the haplotypes in the GBWT index.

Top-level chains (weakly connected components in the graph) are assigned to a number of jobs that can be later used as GBWT construction jobs. Multiple jobs are run in parallel using OpenMP threads.

Each top-level chain is partitioned into subchains that consist of one or more snarls. Multiple snarls are combined into the same subchain if the minimum distance over the subchain is at most the target length and there are GBWT haplotypes that cross the subchain. We also keep extending the subchain if a haplotype would cross the end in both directions. By doing this, we can avoid sequence loss with haplotypes reversing their direction, while keeping kmers specific to each subchain.

If there are no snarls in a top-level chain, it is represented as a single subchain without boundary nodes.

Haplotypes crossing each subchain are represented using minimizers with a single occurrence in the graph.

Throws std::runtime_error on error in single-threaded parts and exits with std::exit(EXIT_FAILURE) in multi-threaded parts.

unique_minimizers() [1/2]

std::vector< HaplotypePartitioner::kmer_type > vg::HaplotypePartitioner::unique_minimizers ( gbwt::size_type  sequence_id ) const

private

unique_minimizers() [2/2]

std::vector< HaplotypePartitioner::kmer_type > vg::HaplotypePartitioner::unique_minimizers ( sequence_type  sequence, Subchain  subchain  ) const

private

Member Data Documentation

APPROXIMATE_JOBS

constexpr size_t vg::HaplotypePartitioner::APPROXIMATE_JOBS = 32

staticconstexpr

Approximate number of construction jobs to be created.

distance_index

const SnarlDistanceIndex& vg::HaplotypePartitioner::distance_index

gbz

const gbwtgraph::GBZ& vg::HaplotypePartitioner::gbz

minimizer_index

const minimizer_index_type& vg::HaplotypePartitioner::minimizer_index

r_index

const gbwt::FastLocate& vg::HaplotypePartitioner::r_index

SUBCHAIN_LENGTH

constexpr size_t vg::HaplotypePartitioner::SUBCHAIN_LENGTH = 10000

staticconstexpr

Target length of a subchain.

verbosity

Verbosity vg::HaplotypePartitioner::verbosity

---

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

• src/recombinator.hpp
• src/recombinator.cpp