Maximum Likelihood Estimation of the Pedigree

Share:

Description

Finds the MLE pedigree using the genetic data only. An approximation is used for genotyping error.

Usage

1
2
MLE.ped(X.list, ped=NULL, USdam=FALSE, nUSdam=NULL, USsire=FALSE, 
   nUSsire=NULL, threshold=0, checkP)

Arguments

X.list

list of design matrices for each offspring derived using getXlist

ped

optional pedigree with id, dam and sire in ech column

USdam

logical or character; if TRUE a single undiferentiated population of unsampled females exists. If USdam is a character vector it must have the same length as id with factor levels representing sub-populations (in time or space) over which the number of unsampled females vary.

nUSdam

numeric vector for number of unsampled females

USsire

logical or character; if TRUE a single undiferentiated population of unsampled males exists. If USsire is a character vector it must have the same length as id with factor levels representing sub-populations (in time or space) over which the number of unsampled males vary.

nUSsire

numeric vector for number of unsampled males

threshold

threshold probability under which ML parents are replaced by NA

checkP

if TRUE the pedigree is checked for legality, and illegal pedigrees rejected. If FALSE it is assumed that any potential parent would produce a legal pedigree, i.e one without circuits, in the terminology of graph theory. Legality is checked

Details

ML estimation of the pedigree is based on the Mendelian transition probabilities in the presence of genotyping error as outlined in Kalinwoski (2006). The probability that the ML parents are the true parents is simply the Mendelian transition probability for those parents divided by the sum of the transition probabilities for the remaining potential parents, both sampled and unsampled. If ped exists and the dam column contains known dam assignemnts and the sire column contains only NA's, then the ML sires will be returned conditional on the dam assignements being true. ML dam estimation with known sires can be performed in the same way. Individuals whose parents cannot be assigned with the required level of certainty (threshold), or whose parents belong to the base or unsampled population, have NA in the dam and sire columns. If each indiviual's potential parents are such that an illegal pedigree could be sampled then checkP=TRUE can be used to ensure legality. This is recommended if the pedigree is to be passed as a starting pedigree to MCMCped. It should be noted that under these circumstances it is possible that multiple pedigrees max exist with the same likelihood and this may not be obvious from the MLE.ped output since assignments are made conditional on earlier assignement being true. As an example, if there are two indiviuals both of which could potentially be each others parents then assigning both to be each others parent is illegal (since each indiviual would be its own grandparent). In simple situations, the parent-offspring and offspring-parent assignements have equal probability, but when checkP=TRUE the first indiviual would have zero probability of being the second individual's parent if the second individual was already assigned as the first individual's parent.

Value

P

pedigree with id in the first column, and dam and sire in the second and third columns

prob

probability of the most likely parental combination

Author(s)

Jarrod Hadfield j.hadfield@ed.ac.uk

References

Hadfield J.D. et al (2006) Molecular Ecology 15 3715-31 Marshall J.D. et al (1998) Molecular Ecology 7 639-655 Kalinowski S.T. et al, Molecular Ecology in press

See Also

MCMCped

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
## Not run: 
data(WarblerP)
data(WarblerG)

GdP<-GdataPed(WarblerG)

res1<-expression(varPed("offspring", restrict=0))
res2<-expression(varPed("terr", gender="Female", relational="OFFSPRING", 
  restrict="=="))

PdP<-PdataPed(formula=list(res1,res2), data=WarblerP, USsire=TRUE)

X.list<-getXlist(PdP=PdP, GdP=GdP, E2=0.005)

ped<-MLE.ped(X.list, USsire=TRUE, nUSsire=10, threshold=0.75)

## End(Not run)

Want to suggest features or report bugs for rdrr.io? Use the GitHub issue tracker.