represent sparsified data of RNA ensemble More...

#include <rna_data.hh>

Inheritance diagram for LocARNA::RnaData:

Public Types
typedef SparseMatrix< double >	arc_prob_matrix_t
	arc probability matrix
typedef size_t	size_type
	usual size type
Public Member Functions
	RnaData (const RnaEnsemble &rna_ensemble, double p_bpcut, double max_bps_length_ratio, const PFoldParams &pfoldparams)
	Construct from RnaEnsemble with cutoff probability.
	RnaData (const std::string &filename, double p_bpcut, double max_bps_length_ratio, const PFoldParams &pfoldparams)
	Construct from file.
	RnaData (const RnaData &rna_dataA, const RnaData &rna_dataB, const Alignment &alignment, double p_expA, double p_expB, bool only_local=false)
	Construct as consensus of two aligned RNAs.
virtual	~RnaData ()
	destructor
const Sequence &	sequence () const
	Get the multiple alignment as sequence.
const MultipleAlignment &	multiple_alignment () const
	Get the multiple alignment.
size_type	length () const
	Get the sequence length.
double	arc_cutoff_prob () const
	Get base pair cutoff probability.
double	arc_prob (pos_type i, pos_type j) const
	Get arc probability.
std::string	mea_structure (double gamma=1.) const
	maximum expected accuracy structure
vrna_plist_t *	plist () const
	Construct plist (pair list of Vienna RNA)
double	joint_arc_prob (pos_type i, pos_type j) const
	Get arc probability.
double	stacked_arc_prob (pos_type i, pos_type j) const
	Get arc probability.
double	prob_paired_upstream (pos_type i) const
	Probability that a position is paired upstream.
double	prob_paired_downstream (pos_type i) const
	Probability that a position is paired downstream.
double	prob_unpaired (pos_type i) const
	Unpaired probability.
std::ostream &	write_pp (std::ostream &out, double p_outbpcut=0) const
std::ostream &	write_size_info (std::ostream &out) const
	Write object size information.
bool	has_stacking () const
	Availability of stacking terms.
void	set_anchors (const SequenceAnnotation &anchors)
	Write access to alignment anchors.
Protected Types
typedef arc_prob_matrix_t::const_iterator	arc_probs_const_iterator
	type of constant iterator over arcs with probability above cutoff
Protected Member Functions
	RnaData (double p_bpcut, size_t max_bp_span)
	Almost empty constructor.
arc_probs_const_iterator	arc_probs_begin () const
	begin of arcs with probability above cutoff Supports iteration over arcs
arc_probs_const_iterator	arc_probs_end () const
	begin of arcs with probability above cutoff Supports iteration over arcs
virtual void	init_from_fixed_structure (const RnaStructure &structure, const PFoldParams &pfoldparams)
	initialize from fixed structure
virtual void	init_from_rna_ensemble (const RnaEnsemble &rna_ensemble, const PFoldParams &pfoldparams)
	initialize from rna ensemble
bool	read_autodetect (const std::string &filename, const PFoldParams &pfoldparams)
	read and initialize from file, autodetect format
virtual bool	inloopprobs_ok () const
	check in loop probabilities
virtual void	read_pp (const std::string &filename)
virtual std::istream &	read_pp (std::istream &in)
void	read_old_pp (const std::string &filename)
void	read_ps (const std::string &filename)
Protected Attributes
RnaDataImpl *	pimpl_
Friends
class	RnaDataImpl
class	ExtRnaDataImpl

Detailed Description

represent sparsified data of RNA ensemble

knows sequence, cutoff probability and base pair probabilities greater than the cutoff probability; potentially knows stacking probabilities

Note:: This class knows the cutoff probability *of its data*. This cutoff can be different from the cutoff in classes like BasePairs which defines the structure elements that are considered by algorithms.; the class guarantees that sequences are normalized (uppercase, T->U) even when read in unnormalized form, e.g. from file or stream

Constructor & Destructor Documentation

LocARNA::RnaData::RnaData	(	const RnaEnsemble &	rna_ensemble,
		double	p_bpcut,
		double	max_bps_length_ratio,
		const PFoldParams &	pfoldparams
	)

Construct from RnaEnsemble with cutoff probability.

Parameters:

rna_ensemble	RNA ensemble data
p_bpcut	cutoff probability
max_bps_length_ratio	max ratio of bps to length (0=no effect)
pfoldparams	folding parameters (controls stacking)

Note:: RnaData copies all required data from rna_ensemble and does not keep a reference; if pfoldparams.stacking is true, copy stacking terms

LocARNA::RnaData::RnaData	(	const std::string &	filename,
		double	p_bpcut,
		double	max_bps_length_ratio,
		const PFoldParams &	pfoldparams
	)

Construct from file.

Parameters:

filename	input file name
p_bpcut	cutoff probability
pfoldparams	folding parameters
max_bps_length_ratio	maximal ratio of number of base pairs divided by sequence length. This serves as a second filter on the "significant" base pairs. Value 0 turns this filter off.

Note:: autodetect format of input; for fa or aln input formats, predict base pair probabilities

Todo:: consider to allow reading from istream; use istream::seekg(0) to reset stream to beginning (needed for format autodetect.) Is there a problem due to fail on eofbit in C++98?

Todo:: filter by maxBPspan when reading base pair probabilities from file. currently maxBPspan in pfoldparams is only respected when folding

Note:: If probabilities have to be computed by folding the input sequence(s), folding is subject to the parameters pfoldparams. Also when reading ensemble probabilities from file, the outcome depends on settings in pfoldparams (stacking, max_bp_span). Without stacking, stacking probs are ignored and max_bp_span is used to filter the base pairs by their maximum span.

LocARNA::RnaData::RnaData	(	const RnaData &	rna_dataA,
		const RnaData &	rna_dataB,
		const Alignment &	alignment,
		double	p_expA,
		double	p_expB,
		bool	only_local = `false`
	)

Construct as consensus of two aligned RNAs.

Parameters:

rna_dataA	RNA ensemble data A
rna_dataB	RNA ensemble data B
alignment	Alignment of A and B
p_expA	background probability for A
p_expB	background probability for B
only_local	if true, construct only local alignment

The object uses mean cutoff probability of the given objects; The background probability is used in computing consensus probabilities. If both input rna data objects have stacking probabilities, stacking consensus probabilities are computed as well. If the object contain sequence anchors, we construct the new object with a consensus anchor string. (The latter is done as part of the consensus sequence computation.)

LocARNA::RnaData::RnaData	(	double	p_bpcut,
		size_t	max_bp_span
	)		`[explicit, protected]`

Almost empty constructor.

Parameters:

p_bpcut	cutoff probability
max_bp_span	maximum base pair span

Member Function Documentation

double LocARNA::RnaData::arc_cutoff_prob ( ) const

Get base pair cutoff probability.

Returns:: cutoff probability p_bpcut

double LocARNA::RnaData::arc_prob	(	pos_type	i,
		pos_type	j
	)		const

Get arc probability.

Parameters:

i	left sequence position
j	right sequence position

Returns:: probability p_ij of basepair (i,j) if p_ij>p_bpcut; otherwise, 0

RnaData::arc_probs_const_iterator LocARNA::RnaData::arc_probs_begin ( ) const [protected]

begin of arcs with probability above cutoff Supports iteration over arcs

Returns:: constant iterator

RnaData::arc_probs_const_iterator LocARNA::RnaData::arc_probs_end ( ) const [protected]

begin of arcs with probability above cutoff Supports iteration over arcs

Returns:: constant iterator

bool LocARNA::RnaData::has_stacking ( ) const

Availability of stacking terms.

Returns:: whether stacking terms are available

void LocARNA::RnaData::init_from_fixed_structure	(	const RnaStructure &	structure,
		const PFoldParams &	pfoldparams
	)		`[protected, virtual]`

initialize from fixed structure

Parameters:

structure	fixed structure
pfoldparams	folding parameters stacking: whether to initialize stacking terms

Note:: can be overloaded to initialize with additional information (in loop probabilities)

Reimplemented in LocARNA::ExtRnaData.

void LocARNA::RnaData::init_from_rna_ensemble	(	const RnaEnsemble &	rna_ensemble,
		const PFoldParams &	pfoldparams
	)		`[protected, virtual]`

initialize from rna ensemble

Parameters:

rna_ensemble	rna ensemble
pfoldparams	folding parameters stacking: whether to initialize stacking terms

Note:: can be overloaded to initialize with additional information (in loop probabilities); this method *never* removes lonely or too long base pairs (according to noLP or maxBPspan, resp.)

Reimplemented in LocARNA::ExtRnaData.

virtual bool LocARNA::RnaData::inloopprobs_ok ( ) const [inline, protected, virtual]

check in loop probabilities

Returns:: true iff loop probabilities are available or not required

Note:: use to indicate the need for recomputation in read_autodetect(); always true in RnaData

Reimplemented in LocARNA::ExtRnaData.

double LocARNA::RnaData::joint_arc_prob	(	pos_type	i,
		pos_type	j
	)		const

Get arc probability.

Parameters:

i	left sequence position
j	right sequence position

Returns:: joint probability p^(2)_ij of basepair (i,j) and (i+1,j-1) if p_i+1j-1>p_bpcut and p^(2)_ij > p_bpcut; otherwise, 0

size_type LocARNA::RnaData::length ( ) const

Get the sequence length.

Returns:: length of RNA sequence

std::string LocARNA::RnaData::mea_structure ( double gamma = 1. ) const

maximum expected accuracy structure

Parameters:

gamma

Returns:: maximum non-crossing expected accuracy structure

Works as interface to the RNAlib function MEA. From the ViennaRNA docu: Each base pair (i,j) gets a score 2*gamma*p_ij and the score of an unpaired base is given by the probability of not forming a pair (compare RNAfold)

const MultipleAlignment & LocARNA::RnaData::multiple_alignment ( ) const

Get the multiple alignment.

Returns:: multiple alignment

vrna_plist_t * LocARNA::RnaData::plist ( ) const

Construct plist (pair list of Vienna RNA)

Note:: the plist has to be deleted by the caller

Returns:: pointer to plist

double LocARNA::RnaData::prob_paired_downstream ( pos_type i ) const

Probability that a position is paired downstream.

Parameters:

i	sequence position

Returns:: probability that a position i is paired with a position j<i (downstream)

Note:: O(sequence.length()) implementation

See also:: prob_paired_upstream

double LocARNA::RnaData::prob_paired_upstream ( pos_type i ) const

Probability that a position is paired upstream.

Parameters:

i	sequence position

Returns:: probability that a position i is paired with a position j>i (upstream)

Note:: O(sequence.length()) implementation

See also:: prob_paired_downstream

double LocARNA::RnaData::prob_unpaired ( pos_type i ) const

Unpaired probability.

Parameters:

i	sequence position

Returns:: probability that a position i is unpaired

Note:: O(sequence.length()) implementation

bool LocARNA::RnaData::read_autodetect	(	const std::string &	filename,
		const PFoldParams &	pfoldparams
	)		`[protected]`

read and initialize from file, autodetect format

Parameters:

filename	name of input file
pfoldparams	folding parameters stacking: whether to initialize stacking terms max_bp_span: maximum base pair span

Returns:: whether probabilities were read completely

Note:: : this method is designed such that it can be used for RnaData and ExtRnaData; the method delegates actual reading to methods read_pp(), read_old_pp(), read_ps(), and the MultipleAlignment class.; when reading in, base pairs exceeding max_bp_span_ or structure information below the probability thresholds are ignored (which is -in part- job of the delegates).

void LocARNA::RnaData::read_old_pp ( const std::string & filename ) [protected]

Read data in the old pp format

Parameters:

filename name of input file

Reads only base pairs with probabilities greater than p_bpcut_; reads stacking probabilities only if has_stacking_ is true

Note:: the old pp format starts with the sequence/alignment and then simply lists the arcs (i,j) with their probabilities p and optionally stacking probabilities p2. pp-files contain entries i j p [p2]. p denotes the probabilitiy of base pair (i,j). The optional stacking probability p2 is the joint probability of base pairs (i,j) and (i+1,j+1).; handling of stacking: after the call, has_stacking_ is true only if the file specified at least one stacking probability and has_stacking_ was true before.

Todo:: move implementation to impl class

void LocARNA::RnaData::read_pp ( const std::string & filename ) [protected, virtual]

Read data in pp format 2.0

Parameters:

filename name of input file

Reads only base pairs with probabilities greater than p_bpcut_; reads stacking probabilities only if has_stacking_ is true

Note:: can be overloaded to read extension sections; pp is a proprietary format of LocARNA. In its simplest version, it starts with the sequence/alignment and then simply lists the arcs (i,j) with their probabilities p and optionally stacking probabilities p2. pp-files contain entries i j p [p2]. p denotes the probabilitiy of base pair (i,j). The optional stacking probability p2 is the joint probability of base pairs (i,j) and (i+1,j+1).; handling of stacking: after the call, has_stacking_ is true only if the file specified the STACK keyword and has_stacking_ was true before.

std::istream & LocARNA::RnaData::read_pp ( std::istream & in ) [protected, virtual]

Read data in pp format 2.0

Parameters:

in	input stream

See also:: read_pp(std::string)

Reimplemented in LocARNA::ExtRnaData.

void LocARNA::RnaData::read_ps ( const std::string & filename ) [protected]

Read data in Vienna's dot plot ps format

Parameters:

filename name of input file

Reads only base pairs with probabilities greater than p_bpcut_; reads stacking probabilities only if has_stacking_ is true

Note:: reads sequence name from file (instead of guessing from filename!); stacking probabilities are read if available (then, sets has_stacking_ to true); throws wrong_format exception if not in ps format

Todo:: move implementation to impl class

sequence characters should be upper case, and Ts translated to Us

const Sequence & LocARNA::RnaData::sequence ( ) const

Get the multiple alignment as sequence.

Returns:: sequence

void LocARNA::RnaData::set_anchors ( const SequenceAnnotation & anchors )

Write access to alignment anchors.

Parameters:

anchors alignment anchors

See also:: MultipleAlignment::set_annotation()

double LocARNA::RnaData::stacked_arc_prob	(	pos_type	i,
		pos_type	j
	)		const

Get arc probability.

Parameters:

i	left sequence position
j	right sequence position

Returns:: conditional probability p_ij|i+1j-1 of basepair (i,j) under condition of base pair (i+1,j-1) if p_i+1j-1>p_bpcut and p^(2)_ij > p_bpcut; throw exception if p_i+1j-1<=p_bpcut; otherwise, 0

std::ostream & LocARNA::RnaData::write_pp	(	std::ostream &	out,
		double	p_outbpcut = `0`
	)		const

Write data in pp format

Parameters:

out	output stream
p_outbpcut	cutoff probability

Returns:: stream

Writes only base pairs with probabilities greater than p_outbpcut

std::ostream & LocARNA::RnaData::write_size_info ( std::ostream & out ) const

Write object size information.

Parameters:

out	output stream

Writes numbers of stored probabilities to stream

Reimplemented in LocARNA::ExtRnaData.

Member Data Documentation

RnaDataImpl* LocARNA::RnaData::pimpl_ [protected]

pointer to corresponding implementation object

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

src/LocARNA/rna_data.hh
src/LocARNA/rna_data.cc

Public Types

Public Member Functions

Protected Types

Protected Member Functions

Protected Attributes

Friends

Detailed Description

Constructor & Destructor Documentation

Member Function Documentation

Member Data Documentation