Sequence::Recombination Namespace Reference
[Analysis of molecular population genetic data]

Methods dealing with recombination. More...

Functions
double	HudsonsC (const Sequence::PolyTable *data, const bool &haveOutgroup, const unsigned &outgroup)
std::vector< std::vector < double > >	Disequilibrium (const Sequence::PolyTable *data, const bool &haveOutgroup, const unsigned &outgroup, const unsigned &mincount, const double max_distance)
bool	Disequilibrium (const Sequence::PolyTable *data, vector< double > &return_values, unsigned *i, unsigned *j, const bool &haveOutgroup, const unsigned &outgroup, const unsigned &mincount, const double max_distance)
bool	Disequilibrium (const Sequence::PolyTable *data, std::vector< double > &return_values, unsigned *i, unsigned *j, const bool &haveOutgroup=false, const unsigned &outgroup=0, const unsigned &mincount=1, const double max_distance=std::numeric_limits< double >::max())
Variables
const double	PRECISION = FLT_EPSILON
const double	CMAX = 10000

---

Detailed Description

Methods dealing with recombination.

This namespace exists primarily so that the file Poly.cc (which defines Sequence::Poly) does not get too large. The routines defined in this namespace all have to do with properties of the association between sites. Current methods implemented are:
1.) Recombination::HudsonsC, which calculates Hudson's C, aka

2.) Recombination::Disequilibrium, which calculated several measures of LD for all pairs of sites, and implements a frequency filter to remove low-frequency variants if desired.

---

Function Documentation

bool Sequence::Recombination::Disequilibrium	(	const Sequence::PolyTable *	data,
		vector< double > &	return_values,
		unsigned *	i,
		unsigned *	j,
		const bool &	haveOutgroup,
		const unsigned &	outgroup,
		const unsigned &	mincount,
		const double	max_distance
	)

More efficient calculation of LD stats

Parameters:

	data	The polymorphism data
	return_values	a vector with enough space to store 6 elements
	i	a pointer to site i. Should be initialized to 0 on the first call
	j	a pointer to site j. Should be initialized to 1 on the first call
	haveOutgroup	true if data contains an outgroup, false otherwise
	outgroup	the index of the outgroup in data haveOutgroup is true
	mincount	the minimum sample frequency to include
	max_distance	max distance between markers (on scale in which positions in data are stored) for inclusion

Returns:

a list of 6 doubles: sitei sitej rsq D D', and the 6th element is 1 if the site pair was skipped (in which case the other 5 elements are meaningless), and 0 otherwise

Definition at line 537 of file Recombination.cc.

double Sequence::Recombination::HudsonsC	(	const Sequence::PolyTable *	data,
		const bool &	haveOutgroup,
		const unsigned &	outgroup
	)

Returns Hudson's (1987) Genetical Research 50:245-250 moment estimator of the population recombination rate. Please note that the properties of this estimator are not ideal, and one should prefer Hudson (2001) Genetics 159: 1805-1817 or McVean et al. (2002) Genetics 160: 1231-1241

---

Variable Documentation

const double Sequence::Recombination::CMAX = 10000

The max value that can be taken by

Definition at line 49 of file Recombination.cc.

const double Sequence::Recombination::PRECISION = FLT_EPSILON

The degree of numeric precision desired in calculating FLT_EPSILON from cfloat is used, which has a precision of about 10^-7 on my machine. I tried DBL_EPSILON, but the numerical routines never seemed to finish...

Definition at line 45 of file Recombination.cc.