rforensicbatwing (version 1.3.1)

Inference from STR data using coalescent theory: Inference from STR data using coalescent theory (e.g. calculating forensic trace-suspect match probability)

Description

Estimates the trace-suspect match probability for a lineage haplotype of STR markers using coalescent theory.

Usage

coalmatchprob(database, haplotype, 
  reps = 10, burnin = 0, treebetN = 10, Nbetsamp = 10, 
  muprior = "constant(0.003)", Nprior = "lognormal(9, 1)", alphaprior = NULL, 
  progress = TRUE, trace = FALSE)
batwing(database, 
  reps = 10, burnin = 0, treebetN = 10, Nbetsamp = 10, 
  muprior = "constant(0.003)", Nprior = "lognormal(9, 1)", alphaprior = NULL, 
  progress = TRUE, trace = FALSE) 
# S3 method for batwing
print(x, ...)
# S3 method for forensicbatwing
plot(x, ...)

Arguments

database

Reference STR database.

haplotype

Haplotype of the suspect.

reps

Number of output lines.

burnin

Number of reps to take before starting recording data.

treebetN

The number of times that changes to the genealogical tree are attempted before any changes to the hyperparameters are attempted. Thus BATWING outputs are separated by treebetN * Nbetsamp attempted tree updates.

Nbetsamp

The number of times that changes to hyperparameters are attempted between outputs.

muprior

Either a single prior distribution for the mutation rate or a vector of prior distributions (one for each locus). If only one prior is supplied, the same mutation rate is used for all loci. If one prior per locus is supplied, each locus has its own chain of mutation rates.

Nprior

Prior distribution of the effective population size.

alphaprior

If NULL, there is no growth (constant population size). If a prior distribution is specified, this gives exponential growth at rate alpha at all times.

progress

Whether to print progress or not.

trace

Whether to print extra trace information or not.

x

A batwing or forensicbatwing object.

...

Not used

Value

coalmatchprob

An object of type forensicbatwing and batwing (for now, forensicbatwing just means that there is a p column in the result matrix).

batwing

An object of type batwing

batwing-object

parameters

The parameters used

priors

The priors used

result

Matrix with information about the iterations (forensic match probability (if coalmatchprob), current value of N, mu's, tree height T, total branch length L, population growth alpha, and loglikelihoods for times, mutations and priors).

proposals_tree

The number of new tree proposals.

accepted_tree

The number of new trees accepted.

proposals_hyperparameters

The number of new hyperparameters proposals.

accepted_hyperparameters

The number of new hyperparameters accepted.

Details

Note that the batwing function runs a standard coalescent inference as described in I.J. Wilson (1999, 2003).

Note that, in contrast to the original BATWING program, migration is not supported. Neither is BATWING's sizemodel=2 (constant-sized population up to a time from where there is exponential growth).

Valid prior distributions:

• uniform(v1, v2) uniform on the interval (v1, v2).

• constant(v1) constant value v1.

• normal(v1, v2) Normal distribution with mean = v1 and sd = v2.

• lognormal(v1, v2) If X has this distribution then log(X) has the normal(v1, v2) distribution.

• gamma(v1, v2) Gamma distributinon with shape v1 and rate v2 giving mode = (v1-1)/v2 and mean = v1/v2.

• beta(v1, v2) Beta distribution with shape parameters v1 and v2 giving mean = v1/(v1 + v2) and the variance is (v1*v2)/((v1 + v2)^2 * (v1 + v2 + 1))

References

The BATWING manual

I.J. Wilson, D.J. Balding, Genealogical inference from microsatellite data, Genetics 150 (1998) 499-510.

I.J. Wilson, M.E. Weale, D.J. Balding, Inferences from DNA data: population histories, evolutionary processes and forensic match probabilities, J. R. Stat. Soc. Ser. A 166 (2003) 155-201.

M.M. Andersen, A. Caliebe, A. Jochens, S. Willuweit, M. Krawczak, Estimating trace-suspect match probabilities for singleton Y-STR haplotypes using coalescent theory, Forensic Sci. Int. Genet. (In Press, Corrected Proof 10.1016/j.fsigen.2012.11.004.).

Examples

# NOT RUN {
database <- matrix(c(1, 1, 2, 2, 1, 3), 3, 2)
haplotype <- c(1, 1)

# coalmatchprob:
coalmp <- coalmatchprob(database, haplotype, 
 reps = 1000, burnin = 0, treebetN = 10, Nbetsamp = 10, 
 muprior = c("normal(0.003, 0.001)",  "normal(0.005, 0.001)"), 
 Nprior = "lognormal(9, 1)", 
 alphaprior = NULL, 
 progress = TRUE, trace = FALSE)
coalmp

murange <- range(c(coalmp$result$mu1, coalmp$result$mu2))

par(mfrow = c(2, 2))
plot(coalmp)
plot(coalmp$result$N, type = "l", ylab = "N")

plot(coalmp$result$mu1, type = "l", col = "red", ylim = murange, ylab = "mu")
points(coalmp$result$mu2, type = "l", col = "blue")

hist(coalmp$result$mu1, col = "#FF000066", 
  xlim = murange, ylim = c(0, 250), main = NULL, xlab = "mu")
hist(coalmp$result$mu2, add = TRUE, col = "#0000FF66")
par(mfrow = c(1, 1))

# batwing:
bw <- batwing(database, 
 reps = 10000, burnin = 1000, treebetN = 10, Nbetsamp = 10, 
 muprior = "normal(0.003, 0.001)", 
 Nprior = "lognormal(9, 1)", 
 alphaprior = NULL, 
 progress = TRUE, trace = FALSE)
bw
# }