GeneMSA Class Reference

multiple sequence alignment of genomes for comparative gene prediction More...

#include <geneMSA.hh>

Public Member Functions
	GeneMSA (RandSeqAccess *rsa, Alignment *a)
string	getSeqID (int speciesIdx)
Strand	getStrand (int speciesIdx)
int	getStart (int speciesIdx)
int	getEnd (int speciesIdx)
list< ExonCandidate * > *	getExonCands (int speciesIdx)
Alignment *	getAlignment ()
vector< int >	getOffsets ()
map< string, ExonCandidate * > *	getECHash (list< ExonCandidate * > *ec)
void	createExonCands (int s, const char *dna, map< int_fast64_t, ExonCandidate * > &ecs, map< int_fast64_t, ExonCandidate * > &addECs)
void	setExonCands (vector< map< int_fast64_t, ExonCandidate * > > &ecs)
void	createOrthoExons (list< OrthoExon > &orthoExonsList, map< int_fast64_t, list< pair< int, ExonCandidate * > > > &alignedECs, Evo *evo, float consThres=0.0, int minAvLen=0)
void	printStats ()
void	printGeneRanges ()
void	printExonCands ()
void	printOrthoExons (list< OrthoExon > &orthoExonsList)
void	computeOmegas (list< OrthoExon > &orthoExonsList, vector< AnnoSequence * > const &seqRanges, PhyloTree *ctree)
void	computeOmegasEff (list< OrthoExon > &orthoExonsList, vector< AnnoSequence * > const &seqRanges, PhyloTree *ctree, ofstream *codonAli)
void	computeClamsaEff (list< OrthoExon > &orthoExonsList, vector< AnnoSequence * > const &seqRanges, PhyloTree *ctree, ofstream *codonAli)
void	computeClamsa (list< OrthoExon > &orthoExonsList, vector< AnnoSequence * > const &seqRanges, PhyloTree *ctree, ofstream *codonAli)
vector< string >	pruneToBV (vector< string > *cs, bit_vector bv)
vector< int >	pruneToBV (vector< int > *rfc, bit_vector bv)
double	omegaForCodonTuple (vector< double > *loglik)
void	printOmegaForCodon (string outdir)
void	printCumOmega ()
void	comparativeSignalScoring (list< OrthoExon > &orthoExonsList)
LocusTree *	constructTree ()
void	calcConsScore (list< OrthoExon > &orthoExonsList, vector< AnnoSequence * > const &seqRanges, string outdir)
double	calcColumnScore (int a, int c, int t, int g)
void	consToWig (vector< double > &consScore, string outdir)
void	printSingleOrthoExon (OrthoExon const &oe, bool files=true)
void	collect_features (int species, list< OrthoExon > *hects, SpeciesGraph *speciesgraph)
void	getAllOEMsas (int species, list< OrthoExon > *hects, unordered_map< string, int > *ref_class, vector< AnnoSequence * > const &seqRanges)
StringAlignment	getMsa (OrthoExon const &oe, vector< AnnoSequence * > const &seqRanges, size_t flanking=0)

Static Public Member Functions
static void	setTree (PhyloTree *t)
static void	setCodonEvo (CodonEvo *c)
static void	setCodonEvoDiscr (CodonEvoDiscr *c)
static int	numSpecies ()
static void	openOutputFiles (string outdir)
static void	closeOutputFiles ()

Static Public Attributes
static int	padding = 1000
static int	orthoExonID = 1
static int	geneRangeID = 1
static vector< int >	exonCandID
static unordered_map< bit_vector, PhyloTree *, boost::hash< bit_vector > >	topologies
static vector< ofstream * >	exonCands_outfiles
static vector< ofstream * >	orthoExons_outfiles
static vector< ofstream * >	geneRanges_outfiles_bed
static vector< ofstream * >	geneRanges_outfiles_gff
static vector< ofstream * >	omega_outfiles
static map< vector< string >, pair< vector< double >, int > >	computedCumValues

Detailed Description

multiple sequence alignment of genomes for comparative gene prediction

Author

Mario Stanke

Alexander Gebauer

Constructor & Destructor Documentation

GeneMSA()

GeneMSA::GeneMSA	(	RandSeqAccess *	rsa,
		Alignment *	a
	)

construct the gene ranges now: simple copy. TODO: extend region when apparently part of the alignment is missing human ***********—******************* mouse *******----—******************- chicken ***********-------------------— ^ | extend range here

Member Function Documentation

computeClamsaEff()

void GeneMSA::computeClamsaEff	(	list< OrthoExon > &	orthoExonsList,
		vector< AnnoSequence * > const &	seqRanges,
		PhyloTree *	ctree,
		ofstream *	codonAli
	)

computeClamsaEff

Parameters

[in]	orthoExonsList	all orthoExons in gene range
[in]	seqRange	the sequences that were aligned
[in]	ctree	tree scaled for units of codon substitutions

See also

computeOmegasEff

createOrthoExons()

void GeneMSA::createOrthoExons	(	list< OrthoExon > &	orthoExonsList,
		map< int_fast64_t, list< pair< int, ExonCandidate * > > > &	alignedECs,
		Evo *	evo,
		float	consThres = 0.0,
		int	minAvLen = 0
	)

find all ortholog exon candidates, that are present in at least max(2, consThresh * m) where m <= numSpecies is the number of species that are present in the alignment Only report OrthoExons oe with at least 'minAvLen' as average length of the exon candidates in oe. ortholog exon candidates:

• both splice sites align exactly
• the exon candidate types agrees (single, rsingle, internal0, ...)
• the phases at both boundaries agree (i.e. exon candidate types and length modulo 3) EC coordinates are region-based, as they are used in the OrthoGraph

createOrthoExons

getAllOEMsas()

void GeneMSA::getAllOEMsas	(	int	species,
		list< OrthoExon > *	hects,
		unordered_map< string, int > *	ref_class,
		vector< AnnoSequence * > const &	seqRanges
	)

getAllOEMsas obtains and prints multiple sequence alignments (MSAs) and their label y=0,1, whether it constitutes a real CDS or not in the reference species.

Parameters

[in]	species	is the reference species index
[in]	hects	ortho exons to get alignments from
[in]	ref_class	assignment of CDS keys as 1 (= in reference anno) or 0.
[in]	seqRanges	the aligned sequences

getMsa()

StringAlignment GeneMSA::getMsa	(	OrthoExon const &	oe,
		vector< AnnoSequence * > const &	seqRanges,
		size_t	flanking = 0
	)

getMsa obtains the alignment in string format, including unaligned insertions that are not part of any fragment.

Returns

msa vector of alignment rows

Parameters

[in]	oe	the OrthoExon whose alignment is sought
[in]	seqRanges	contains the alignment of the larger region tuple
[in]	flanking	allows to add padding on both sides, e.g. for sequence signals

---

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

• include/geneMSA.hh
• src/geneMSA.cc