LocARNA::RnaDataImpl Class Reference

Implementation of RnaData. More...

#include <rna_data_impl.hh>

List of all members.

Classes
class	keyvec
Public Types
typedef RnaData::arc_prob_matrix_t	arc_prob_matrix_t
	type for matrix of arc probabilities
Public Member Functions
	RnaDataImpl (RnaData *self, const RnaData &rna_dataA, const RnaData &rna_dataB, const Alignment::edges_t &alignment, double p_expA, double p_expB)
	Construct as consensus of two aligned RNAs.
	RnaDataImpl (RnaData *self, double p_bpcut, size_t max_bp_span)
	Almost empty constructor.
void	init_from_fixed_structure (const RnaStructure &structure, const PFoldParams &pfoldparams)
	initialize from fixed structure
void	init_from_rna_ensemble (const RnaEnsemble &rna_ensemble, const PFoldParams &pfoldparams)
	initialize from rna ensemble
std::istream &	read_pp_sequence (std::istream &in)
	read sequence section of pp-format
std::istream &	read_pp_arc_probabilities (std::istream &in)
	read section of base pair probabilities of pp-format
std::ostream &	write_pp_sequence (std::ostream &out) const
	write section of base pair probabilities of pp-format
std::ostream &	write_pp_arc_probabilities (std::ostream &out, double p_outbpcut, bool stacking) const
	write section of base pair probabilities of pp-format
void	init_as_consensus_dot_plot (const Alignment::edges_t &edges, const RnaData &rna_dataA, const RnaData &rna_dataB, double p_expA, double p_expB, double f_penalty, bool stacking)
	Initialize as consensus of two aligned RNAs.
double	consensus_probability (double pA, double pB, size_t sizeA, size_t sizeB, double p_expA, double p_expB, double p_penalty) const
	Consensus probability.
void	drop_worst_bps (size_t keep)
	Drop base pairs with lowest probability.
Public Attributes
RnaData *	self_
MultipleAlignment	sequence_
	the sequence
double	p_bpcut_
	cutoff probabilitiy for base pair
size_t	max_bp_span_
arc_prob_matrix_t	arc_probs_
arc_prob_matrix_t	arc_2_probs_
bool	has_stacking_
	whether stacking probabilities are available

---

Detailed Description

Implementation of RnaData.

---

Constructor & Destructor Documentation

LocARNA::RnaDataImpl::RnaDataImpl	(	RnaData *	self,
		const RnaData &	rna_dataA,
		const RnaData &	rna_dataB,
		const Alignment::edges_t &	alignment,
		double	p_expA,
		double	p_expB
	)

Construct as consensus of two aligned RNAs.

Parameters:

self	pointer to corresponding RnaData object
rna_dataA	data of RNA A
rna_dataB	data of RNA B
alignment	pairwise alignment of A and B
p_expA	background probability A
p_expB	background probability B

<

Todo:

this constant should be configurable

LocARNA::RnaDataImpl::RnaDataImpl	(	RnaData *	self,
		double	p_bpcut,
		size_t	max_bp_span
	)

Almost empty constructor.

Parameters:

self	pointer to corresponding RnaData object
p_bpcut	cutoff probability

---

Member Function Documentation

double LocARNA::RnaDataImpl::consensus_probability	(	double	pA,
		double	pB,
		size_t	sizeA,
		size_t	sizeB,
		double	p_expA,
		double	p_expB,
		double	p_penalty
	)		const

Consensus probability.

Parameters:

pA	probability A
pB	probability B
sizeA	number of rows in sequence A
sizeB	number of rows in sequence B
p_expA	background probability A
p_expB	background probability B
p_penalty	penalty probability for base pairs with probability below cutoff

Note:

Essentially the consensus base pair probabilities are geometric means of the combined single base pair probabilties. This is maintained by weighting the pairwise consensus computations with the respective numbers of sequences.

The penalty probability p_penalty has to be chosen carefully: Assuming 0 probability for base pairs below of the threshold geometric mean would result in 0 as consensus probability; thus we assume p_penalty in this case, however too high p_penalty leads to the accumulation of many base pairs with small probabilities.

Reasonably, p_penalty must be a small fraction of p_cutoff_.

Returns:

consensus probability

void LocARNA::RnaDataImpl::drop_worst_bps

(

size_t

keep

)

Drop base pairs with lowest probability.

Parameters:

keep	the maximum number of base pairs to keep

void LocARNA::RnaDataImpl::init_as_consensus_dot_plot	(	const Alignment::edges_t &	edges,
		const RnaData &	rna_dataA,
		const RnaData &	rna_dataB,
		double	p_expA,
		double	p_expB,
		double	f_penalty,
		bool	stacking
	)

Initialize as consensus of two aligned RNAs.

Parameters:

edges	alignment edges
rna_dataA	rna data A
rna_dataB	rna data B
p_expA	background probability A
p_expB	background probability B
f_penalty	factor for penalty probability: p_penalty = p_bpcut*f_penalty,

See also:

consensus_probability()

Parameters:

stacking

if true, stacking consensus is computed

void LocARNA::RnaDataImpl::init_from_fixed_structure	(	const RnaStructure &	structure,
		const PFoldParams &	pfoldparams
	)

initialize from fixed structure

Parameters:

structure	fixed structure
pfoldparams	pfold parameters; only stacking is used

Note:

this strictly sets the probability of all base pairs in structure to 1.0. It does not care for constraints like noLP or maxBPspan; this has to be handled by the caller!

void LocARNA::RnaDataImpl::init_from_rna_ensemble	(	const RnaEnsemble &	rna_ensemble,
		const PFoldParams &	pfoldparams
	)

initialize from rna ensemble

Parameters:

rna_ensemble	rna ensemble
pfoldparams	folding parameters. if stacking, initialize stacking terms; if noLP, drop lonely pairs

std::istream & LocARNA::RnaDataImpl::read_pp_arc_probabilities

(

std::istream &

in

)

read section of base pair probabilities of pp-format

Parameters:

in	input stream

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

std::istream & LocARNA::RnaDataImpl::read_pp_sequence

(

std::istream &

in

)

read sequence section of pp-format

Parameters:

in	input stream

Returns:

stream

this section comprises sequence/multiple alignment (possibly including sequence anchor annotation)

std::ostream & LocARNA::RnaDataImpl::write_pp_arc_probabilities	(	std::ostream &	out,
		double	p_outbpcut,
		bool	stacking
	)		const

write section of base pair probabilities of pp-format

Write arc probabilities.

Parameters:

out	ouput stream
p_outbpcut	cutoff probabilitiy
stacking	whether to write stacking probabilities; if stacking but !has_stacking_, no stacking terms are written but flag #STACKS is written to output

Returns:

stream

Write only base pairs with probabilities greater than p_outbpcut

Writes arc and stacking probabilities to stream; filters by probability threshold p_outbpcut

std::ostream & LocARNA::RnaDataImpl::write_pp_sequence

(

std::ostream &

out

)

const

write section of base pair probabilities of pp-format

Parameters:

out	ouput stream

Returns:

stream

---

Member Data Documentation

arc_prob_matrix_t LocARNA::RnaDataImpl::arc_2_probs_

sparse array for all probabilities that a pair (i,j) and its immediately inner pair (i+1,j-1) are formed simultaneously above threshold; analogous to arc_probs_

Note:

arc_2_probs_ has entry (i,j) implies arc_probs_ has entry (i,j)

arc_prob_matrix_t LocARNA::RnaDataImpl::arc_probs_

sparse array for all arc probabilities above threshold; the array is used when reading in the probabilities and for merging probs during pp-output

RnaData* LocARNA::RnaDataImpl::self_

• pointer to corresponding non-impl object

---

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

• src/LocARNA/rna_data_impl.hh
• src/LocARNA/rna_data.cc