sample

Class RecombSingleBaum

java.lang.Object

sample.RecombSingleBaum

All Implemented Interfaces:

SingleBaumInterface

public class RecombSingleBaum
extends java.lang.Object
implements SingleBaumInterface

Class RestrictedSingleBaum implements the Baum forward and backward algorithms for a hidden Markov model (HMM) of an individual's genotype data. The HMM transition probabilities model recent genetic recombination by allowing jumps between states that are not connected by a node.

Instances of class RestrictedSingleBaum are not thread-safe.

Constructor Summary

Constructors

Constructor and Description
RecombSingleBaum(SamplerData samplerData, long seed, int nSamplesPerIndividual, boolean lowMem) Creates a new RestrictedSingleBaum instance from the specified data.

Method Summary

Modifier and Type	Method and Description
Dag	dag() Returns the directed acyclic graph that determines the transition probabilities.
GL	gl() Returns the emission probabilities.
int	nSamplesPerIndividual() Returns the number of haplotype pairs that are sampled for each individual.
java.util.List<HapPair>	randomSample(int sample) Returns a list of this.nSamplesPerIndividual() sampled haplotype pairs for the specified individual.
java.util.List<HapPair>	randomSample(int sample, double[] gprobs) Returns a list of this.nSamplesPerIndividual() sampled haplotype pairs for the specified individual.
long	seed() Returns the initial random seed.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

RecombSingleBaum

public RecombSingleBaum(SamplerData samplerData,
                        long seed,
                        int nSamplesPerIndividual,
                        boolean lowMem)

Creates a new RestrictedSingleBaum instance from the specified data.

Parameters:

samplerData - the analysis data

seed - the random seed

nSamplesPerIndividual - the number of haplotype pairs that will be sampled for each individual

lowMem - true if a low memory algorithm should be used, and false otherwise

Throws:

java.lang.IllegalArgumentException - if nSamplesPerIndividual < 1

java.lang.NullPointerException - if samplerData == null

Method Detail

dag

public Dag dag()

Description copied from interface: SingleBaumInterface

Returns the directed acyclic graph that determines the transition probabilities.

Specified by:

dag in interface SingleBaumInterface

Returns:

the directed acyclic graph that determines the transition probabilities

gl

public GL gl()

Description copied from interface: SingleBaumInterface

Returns the emission probabilities.

Specified by:

gl in interface SingleBaumInterface

Returns:

the emission probabilities

nSamplesPerIndividual

public int nSamplesPerIndividual()

Description copied from interface: SingleBaumInterface

Returns the number of haplotype pairs that are sampled for each individual.

Specified by:

nSamplesPerIndividual in interface SingleBaumInterface

Returns:

the number of haplotype pairs that are sampled for each individual

seed

public long seed()

Description copied from interface: SingleBaumInterface

Returns the initial random seed.

Specified by:

seed in interface SingleBaumInterface

Returns:

the initial random seed

randomSample

public java.util.List<HapPair> randomSample(int sample)

Description copied from interface: SingleBaumInterface

Returns a list of this.nSamplesPerIndividual() sampled haplotype pairs for the specified individual. Haplotype pairs are sampled conditional on the HMM with transition probabilities determined by this.dag() and emission probabilities determined by this.gl().

The contract for this method is unspecified if no haplotype pair is consistent with the HMM.

Specified by:

randomSample in interface SingleBaumInterface

Parameters:

sample - a sample index

Returns:

a list of this.nSamplesPerIndividual() sampled haplotype pairs for the specified individual

randomSample

public java.util.List<HapPair> randomSample(int sample,
                                            double[] gprobs)

Description copied from interface: SingleBaumInterface

Returns a list of this.nSamplesPerIndividual() sampled haplotype pairs for the specified individual. Haplotype pairs are sampled conditional on the HMM with transition probabilities determined by this.dag() and emission probabilities determined by this.gl(). Posterior genotype probabilities are written to the specified array. The posterior probability of the j-th genotype for the k-th marker is stored at index gl.markers().sumGenotypes(k) + j in the gtProbs array.

The contract for this method is unspecified if no haplotype pair is consistent with the HMM.

Specified by:

randomSample in interface SingleBaumInterface

Parameters:

sample - the sample index

gprobs - a array to which posterior genotype probabilities for the sample will be written

Returns:

a list of this.nSamplesPerIndividual() sampled haplotype pairs for the specified individual