| Title: | Missing Person Identification Tools |
|---|---|
| Description: | An open source software package written in R statistical language. It consist in a set of decision making tools to conduct missing person searches. Particularly, it allows computing optimal LR threshold for declaring potential matches in DNA-based database search. More recently 'mispitools' incorporates preliminary investigation data based LRs. Statistical weight of different traces of evidence such as biological sex, age and hair color are presented. For citing mispitools please use the following references: Marsico and Caridi, 2023 <doi:10.1016/j.fsigen.2023.102891> and Marsico, Vigeland et al. 2021 <doi:10.1016/j.fsigen.2021.102519>. |
| Authors: | Franco Marsico [aut, cre] |
| Maintainer: | Franco Marsico <[email protected]> |
| License: | GPL (>= 3) |
| Version: | 1.2.0 |
| Built: | 2024-11-08 02:42:44 UTC |
| Source: | https://github.com/marsicofl/mispitools |
STRs allelic frequencies from specified country.
ArgentinaArgentina
A data frame allele frequencies
A dataset of allele frequencies.
AsiaAsia
A data frame allele frequencies
STRs allelic frequencies from specified country.
AustriaAustria
A data frame allele frequencies
This function calculates the Kullback-Leibler divergence for shared genetic markers between two populations, considering allele frequencies. It normalizes data, adjusts zero frequencies, and calculates divergence in both directions.
bidirectionalKL(data1, data2, minFreq = 1e-10)bidirectionalKL(data1, data2, minFreq = 1e-10)
data1 |
DataFrame with allele frequencies for the first population. |
data2 |
DataFrame with allele frequencies for the second population. |
minFreq |
Minimum frequency to be considered for unobserved or poorly observed alleles. |
A list containing the Kullback-Leibler divergence from data1 to data2 and vice versa.
bidirectionalKL(Argentina, BosniaHerz)bidirectionalKL(Argentina, BosniaHerz)
STRs allelic frequencies from specified country.
BosniaHerzBosniaHerz
A data frame allele frequencies
STRs allelic frequencies from specified country.
ChinaChina
A data frame allele frequencies
Epsilon hair color matrix
Cmodel( errorModel = c("custom", "uniform")[1], ep = 0.01, ep12 = 0.01, ep13 = 0.005, ep14 = 0.01, ep15 = 0.003, ep23 = 0.01, ep24 = 0.003, ep25 = 0.01, ep34 = 0.003, ep35 = 0.003, ep45 = 0.01 )Cmodel( errorModel = c("custom", "uniform")[1], ep = 0.01, ep12 = 0.01, ep13 = 0.005, ep14 = 0.01, ep15 = 0.003, ep23 = 0.01, ep24 = 0.003, ep25 = 0.01, ep34 = 0.003, ep35 = 0.003, ep45 = 0.01 )
errorModel |
custom allows selecting a specfic epsilon for each MP-UHR pair, uniform use ep for all. |
ep |
epsilon |
ep12 |
epsilon |
ep13 |
epsilon |
ep14 |
epsilon |
ep15 |
epsilon |
ep23 |
epsilon |
ep24 |
epsilon |
ep25 |
epsilon |
ep34 |
epsilon |
ep35 |
epsilon |
ep45 |
epsilon |
A value of Likelihood ratio based on preliminary investigation data. In this case, sex.
Cmodel()Cmodel()
Combine LRs: a function for combining LRs obtained from simulations.
combLR(LRdatasim1, LRdatasim2)combLR(LRdatasim1, LRdatasim2)
LRdatasim1 |
A data frame object with the results of simulations. Outputs from simLRgen or simLRprelim funcionts. |
LRdatasim2 |
A second data frame object with the results of simulations. Outputs from simLRgen or simLRprelim funcionts. |
An object of class data.frame combining the LRs obtained from simulations (the function multiplies the LRs).
library(mispitools) library(forrel) x = linearPed(2) x = setMarkers(x, locusAttributes = NorwegianFrequencies[1:5]) x = profileSim(x, N = 1, ids = 2) LRdatasim1 = simLRgen(x, missing = 5, 10, 123) LRdatasim2 = simLRprelim("sex") combLR(LRdatasim1,LRdatasim2)library(mispitools) library(forrel) x = linearPed(2) x = setMarkers(x, locusAttributes = NorwegianFrequencies[1:5]) x = profileSim(x, N = 1, ids = 2) LRdatasim1 = simLRgen(x, missing = 5, 10, 123) LRdatasim2 = simLRprelim("sex") combLR(LRdatasim1,LRdatasim2)
This function calculates the Likelihood Ratios (LRs) for each combination of hair colour,
skin colour, and eye colour between two datasets. It assumes one dataset (conditioned)
contains numerators and the other (unconditioned) contains denominators.
compute_LRs_colors(conditioned, unconditioned)compute_LRs_colors(conditioned, unconditioned)
conditioned |
A dataframe with at least the columns 'hair_colour', 'skin_colour', 'eye_colour', and 'numerators'. |
unconditioned |
A dataframe with at least the columns 'hair_colour', 'skin_colour', 'eye_colour', and 'f_h_s_y'. |
A dataframe with the merged data and computed LRs.
data <- simRef() conditioned <- conditionedProp(data, 1, 1, 1, 0.01, 0.01, 0.01) unconditioned <- refProp(data) compute_LRs_colors(conditioned, unconditioned)data <- simRef() conditioned <- conditionedProp(data, 1, 1, 1, 0.01, 0.01, 0.01) unconditioned <- refProp(data) compute_LRs_colors(conditioned, unconditioned)
This function calculates the conditioned proportions for pigmentation traits for UP, when UP is MP. It considers error rates for observations of hair color, skin color, and eye color.
conditionedProp(data, h, s, y, eh, es, ey)conditionedProp(data, h, s, y, eh, es, ey)
data |
A data.frame containing the characteristics of UPs. |
h |
An integer representing the MP's hair color. |
s |
An integer representing the MP's skin color. |
y |
An integer representing the MP's eye color. |
eh |
A numeric value representing the error rate for observing hair color. |
es |
A numeric value representing the error rate for observing skin color. |
ey |
A numeric value representing the error rate for observing eye color. |
A numeric vector containing the conditioned proportion (numerator) for each individual in the dataset. These values are calculated based on the probability of observing the given combination of characteristics in the MP, compared to each UP.
General plot for condiionted probabilities and LR combining variables
CondPlot(CPT_POP, CPT_MP)CondPlot(CPT_POP, CPT_MP)
CPT_POP |
Population conditioned probability table |
CPT_MP |
Missing person conditioned probability table |
A value of Likelihood ratio based on preliminary investigation data. In this case, sex.
Cmodel()Cmodel()
Missing person based conditioned probability
CPT_MP(MPs = "F", MPc = 1, eps = 0.05, epa = 0.05, epc = Cmodel())CPT_MP(MPs = "F", MPc = 1, eps = 0.05, epa = 0.05, epc = Cmodel())
MPs |
Missing person sex |
MPc |
Missing person hair color |
eps |
sex epsilon |
epa |
age epsilon - Age is not specified in this first version, because it asumes uniformity. |
epc |
color model |
A value of Likelihood ratio based on preliminary investigation data. In this case, sex.
CPT_MP()CPT_MP()
Population based conditioned probability
CPT_POP( propS = c(0.5, 0.5), MPa = 40, MPr = 6, propC = c(0.3, 0.2, 0.25, 0.15, 0.1) )CPT_POP( propS = c(0.5, 0.5), MPa = 40, MPr = 6, propC = c(0.3, 0.2, 0.25, 0.15, 0.1) )
propS |
age epsilon - Age is not specified in this first version, because it asumes uniformity. |
MPa |
Missing person sex |
MPr |
Missing person hair color |
propC |
sex epsilon |
A value of Likelihood ratio based on preliminary investigation data. In this case, sex.
CPT_POP()CPT_POP()
Decision making plot: a function for plotting false positive and false negative rates for each LR threshold.
deplot(datasim, LRmax = 1000)deplot(datasim, LRmax = 1000)
datasim |
Input dataframe containing expected LRs for related and unrelated POIs. It should be the output from makeLRsims function. |
LRmax |
Maximum LR value used as a threshold. 1000 setted by default. |
A plot showing false positive and false negative rates for each likelihood ratio threshold.
library(forrel) x = linearPed(2) x = setMarkers(x, locusAttributes = NorwegianFrequencies[1:5]) x = profileSim(x, N = 1, ids = 2) datasim = simLRgen(x, missing = 5, 10, 123) deplot(datasim)library(forrel) x = linearPed(2) x = setMarkers(x, locusAttributes = NorwegianFrequencies[1:5]) x = profileSim(x, N = 1, ids = 2) datasim = simLRgen(x, missing = 5, 10, 123) deplot(datasim)
Decision Threshold: a function for computing likelihood ratio decision threshold.
DeT(datasim, weight)DeT(datasim, weight)
datasim |
Input dataframe containing expected LRs for related and unrelated POIs. It should be the output from makeLRsims function. |
weight |
The differential weight between false positives and false negatives. A value of 10 is suggested. |
A value of Likelihood ratio suggested as threshold based on false positive-false negative trade-off.
library(forrel) x = linearPed(2) x = setMarkers(x, locusAttributes = NorwegianFrequencies[1:5]) x = profileSim(x, N = 1, ids = 2) datasim = simLRgen(x, missing = 5, 10, 123) DeT(datasim, 10)library(forrel) x = linearPed(2) x = setMarkers(x, locusAttributes = NorwegianFrequencies[1:5]) x = profileSim(x, N = 1, ids = 2) datasim = simLRgen(x, missing = 5, 10, 123) DeT(datasim, 10)
STRs allelic frequencies from specified country.
EuropeEurope
A data frame allele frequencies
Function for getting STR allele frequencies from different world populations.
getfreqs(region)getfreqs(region)
region |
select the place of the allele frequency database. Possible values are listed: "Argentina", "Asia", "Europe", "USA", "Austria", "BosniaHerz", "China" and "Japan". |
An allele frequency database adapted compatible with pedtools format.
https://doi.org/10.1016/j.fsigss.2009.08.178; https://doi.org/10.1016/j.fsigen.2016.06.008; https://doi.org/10.1016/j.fsigen.2018.07.013.
STRs allelic frequencies from specified country.
JapanJapan
A data frame allele frequencies
This function computes the Kullback-Leibler (KL) divergence between two probability distributions represented by matrices, using a base 10 logarithm. The function calculates KL divergence in both directions (P || Q and Q || P) and handles zero probabilities by replacing them with a minimum value to avoid undefined logarithms.
klPIE(P, Q, min_value = 1e-12)klPIE(P, Q, min_value = 1e-12)
P |
A numeric matrix representing the first probability distribution. The entire matrix should sum to 1. |
Q |
A numeric matrix representing the second probability distribution. The entire matrix should sum to 1. |
min_value |
A numeric value representing the minimum value to replace
any zero probabilities. Defaults to |
A named numeric vector with two elements:
The KL divergence from P to Q (P || Q).
The KL divergence from Q to P (Q || P).
Likelihood ratio for age variable
LRage( MPa = 40, MPr = 6, UHRr = 1, gam = 0.07, nsims = 1000, epa = 0.05, erRa = epa, H = 1, modelA = c("uniform", "custom")[1], LR = FALSE, seed = 1234 )LRage( MPa = 40, MPr = 6, UHRr = 1, gam = 0.07, nsims = 1000, epa = 0.05, erRa = epa, H = 1, modelA = c("uniform", "custom")[1], LR = FALSE, seed = 1234 )
MPa |
Missing person age |
MPr |
Missing person age range. |
UHRr |
Unidentified person range |
gam |
Simulation parameter for UHR ages. |
nsims |
number of simulations. |
epa |
epsilon age |
erRa |
error rate in the database. |
H |
hipothesis tested, H1: UHR is MP, H2: UHR is not MP. |
modelA |
reference database probabilities, uniform assumes equally probable ages. Custom needs a vector with ages frequencies. |
LR |
compute LR values |
seed |
For reproducible simulations |
A value of Likelihood ratio based on preliminary investigation data. In this case, Age.
Likelihood ratio for color variable
LRcol( MPc = 1, epc = Cmodel(), erRc = epc, nsims = 1000, Pc = c(0.3, 0.2, 0.25, 0.15, 0.1), H = 1, Qprop = MPc, LR = FALSE, seed = 1234 )LRcol( MPc = 1, epc = Cmodel(), erRc = epc, nsims = 1000, Pc = c(0.3, 0.2, 0.25, 0.15, 0.1), H = 1, Qprop = MPc, LR = FALSE, seed = 1234 )
MPc |
MP hair color |
epc |
epsilon paramenter. |
erRc |
error rate in the database. |
nsims |
number of simulations performed. |
Pc |
hair color probabilities. |
H |
hypothesis tested, H1: UHR is MP, H2: UHR is no MP |
Qprop |
Query color tested. |
LR |
compute LR values |
seed |
For reproducible simulations |
A value of Likelihood ratio based on preliminary investigation data. In this case, hair color.
LRcol()LRcol()
Simulate LR values considering H1 and H2
LRcolors(df, seed = 1234, nsim = 500)LRcolors(df, seed = 1234, nsim = 500)
df |
A data.frame containing the characteristics of individuals, numerator, f_h_s_y and LRs. Output from compute_LRs function. |
seed |
For replication purposes. |
nsim |
Number of LRs simulated. |
LR distribution considering H1 (Related) and H2 (Unrelated).
Likelihood ratio for birth date in missing person searches
LRdate( ABD = "1976-05-31", DBD = "1976-07-15", PrelimData, alpha = c(1, 4, 60, 11, 6, 4, 4), cuts = c(-120, -30, 30, 120, 240, 360), draw = 500, type = 1, seed = 123 )LRdate( ABD = "1976-05-31", DBD = "1976-07-15", PrelimData, alpha = c(1, 4, 60, 11, 6, 4, 4), cuts = c(-120, -30, 30, 120, 240, 360), draw = 500, type = 1, seed = 123 )
ABD |
Actual birth date of the missing person. |
DBD |
Declared birth date of the person of interest. |
PrelimData |
Used when type = 2, is the dataframe with the DBD of the persons of interest in the database. |
alpha |
A vector containing the alpha values for the dirichlet. It should contain the number of categories of differences between DBD and ABD. |
cuts |
Value of differences between DBD and ABD used for category definition. |
draw |
Number of simulations for Dirichlet distribution computation. |
type |
Type of scenario, type 1 is an "open search", where it is unknown if the missing person is in the database. Type 2 refers to a scenario where the missing person is in the database. |
seed |
Seed for simulations. |
A value of Likelihood ratio based on preliminary investigation data. In this case, birth date.
library(DirichletReg) LRdate(ABD = "1976-05-31", DBD = "1976-07-15", PrelimData, alpha = c(1, 4, 60, 11, 6, 4, 4), cuts = c(-120, -30, 30, 120, 240, 360), type = 1, seed = 123)library(DirichletReg) LRdate(ABD = "1976-05-31", DBD = "1976-07-15", PrelimData, alpha = c(1, 4, 60, 11, 6, 4, 4), cuts = c(-120, -30, 30, 120, 240, 360), type = 1, seed = 123)
Likelihood ratio distribution: a function for plotting expected log10(LR) distributions under relatedness and unrelatedness.
LRdist(datasim)LRdist(datasim)
datasim |
Input dataframe containing expected LRs for related and unrelated POIs. It should be the output from makeLRsims function. |
A plot showing likelihood ratio distributions under relatedness and unrelatedness hypothesis.
library(forrel) x = linearPed(2) x = setMarkers(x, locusAttributes = NorwegianFrequencies[1:5]) x = profileSim(x, N = 1, ids = 2) datasim = simLRgen(x, missing = 5, 10, 123) LRdist(datasim)library(forrel) x = linearPed(2) x = setMarkers(x, locusAttributes = NorwegianFrequencies[1:5]) x = profileSim(x, N = 1, ids = 2) datasim = simLRgen(x, missing = 5, 10, 123) LRdist(datasim)
Likelihood ratio for sex variable
LRsex( MPs = "F", eps = 0.05, erRs = eps, nsims = 1000, Ps = c(0.5, 0.5), H = 1, LR = FALSE, seed = 1234 )LRsex( MPs = "F", eps = 0.05, erRs = eps, nsims = 1000, Ps = c(0.5, 0.5), H = 1, LR = FALSE, seed = 1234 )
MPs |
MP sex |
eps |
epsilon paramenter. |
erRs |
error rate in the database. |
nsims |
number of simulations performed. |
Ps |
Sex probabilities in the population. |
H |
hypothesis tested, H1: UHR is MP, H2: UHR is no MP |
LR |
compute LR values |
seed |
For reproducible simulations |
A value of Likelihood ratio based on preliminary investigation data. In this case, sex.
LRsex()LRsex()
Make preliminary investigation MP data simulations: a function for obtaining a database of preliminary investigation data for a missing person search.
makeMPprelim( casetype = "children", dateinit = "1975/01/01", scenario = 1, femaleprop = 0.5, ext = 100, numsims = 10000, seed = 123, region = c("North America", "South America", "Africa", "Asia", "Europe", "Oceania"), regionprob = c(0.2, 0.2, 0.2, 0.1, 0.2, 0.1) )makeMPprelim( casetype = "children", dateinit = "1975/01/01", scenario = 1, femaleprop = 0.5, ext = 100, numsims = 10000, seed = 123, region = c("North America", "South America", "Africa", "Asia", "Europe", "Oceania"), regionprob = c(0.2, 0.2, 0.2, 0.1, 0.2, 0.1) )
casetype |
Type of missing person search case. Two options are available: "migrants" or "children". |
dateinit |
Minimun birth date of simulated missing person. Casetype: Children. |
scenario |
Birth date distribution scenarios: (1) non-uniform, (2) uniform. Casetype: Children. |
femaleprop |
Proportion of females. Casetype: All. |
ext |
Time extension for minimun birth date, range in scenario 1 and days in scenario 2. Casetype: Children. |
numsims |
Number of simulated MPs. Casetype: All. |
seed |
Select a seed for simulations. If it is defined, results will be reproducible. Casetype: All. |
region |
Birth region or place in missing children case or place of place of the last seen in missing migrant case. Casetype: All. |
regionprob |
Region proportions. Casetype: All. |
An object of class data.frame with preliminary investigation data.
makeMPprelim()makeMPprelim()
Make POIs gen: a function for obtaining a database with genetic information from simulated POIs or UHRs.
makePOIgen(numsims = 100, reference, seed = 123)makePOIgen(numsims = 100, reference, seed = 123)
numsims |
Number of simulations performed (numer of POIs or UHRs). |
reference |
Indicate the reference STRs/SNPs frequency database used for simulations. |
seed |
Select a seed for simulations. If it is defined, results will be reproducible. Suggested, seed = 123 |
An object of class data.frame with genetic information from POIs (randomly sampled from the frequency database).
library(forrel) freqdata <- getfreqs(Argentina) makePOIgen(numsims = 100, reference = freqdata, seed = 123)library(forrel) freqdata <- getfreqs(Argentina) makePOIgen(numsims = 100, reference = freqdata, seed = 123)
Make preliminary investigation POI/UHR data simulations: a function for obtaining a database of preliminary investigation data for a missing person search.
makePOIprelim( casetype = "children", dateinit = "1975/01/01", scenario = 1, femaleprop = 0.5, ext = 100, numsims = 10000, seed = 123, birthprob = c(0.09, 0.9, 0.01), region = c("North America", "South America", "Africa", "Asia", "Europe", "Oceania"), regionprob = c(0.2, 0.2, 0.2, 0.1, 0.2, 0.1) )makePOIprelim( casetype = "children", dateinit = "1975/01/01", scenario = 1, femaleprop = 0.5, ext = 100, numsims = 10000, seed = 123, birthprob = c(0.09, 0.9, 0.01), region = c("North America", "South America", "Africa", "Asia", "Europe", "Oceania"), regionprob = c(0.2, 0.2, 0.2, 0.1, 0.2, 0.1) )
casetype |
Type of missing person search case. Two options are available: "migrants" or "children". |
dateinit |
Minimun birth date of simulated persons of interest. Casetype: Children. |
scenario |
Birth date distribution scenarios: (1) non-uniform, (2) uniform. Casetype: Children. |
femaleprop |
Proportion of females. Casetype: All. |
ext |
Time extension for minimun birth date, range in scenario 1 and days in scenario 2. Casetype: Children. |
numsims |
Number of simulated POIs/UHRs. Casetype: All. |
seed |
Select a seed for simulations. If it is defined, results will be reproducible. Casetype: All. |
birthprob |
Birth type probabilities: home birth, hospital birth and unknown-adoption. Casetype: Children. |
region |
Birth region or place in missing children case or place of discovery of the human remain in missing migrant case. Casetype: All. |
regionprob |
Region proportions. Casetype: All. |
An object of class data.frame with preliminary investigation data.
makePOIprelim( dateinit = "1975/01/01", scenario = 1, femaleprop = 0.5, ext = 100, numsims = 10000, seed = 123, birthprob = c(0.09, 0.9, 0.01), region = c("North America", "South America", "Africa", "Asia", "Europe", "Oceania"), regionprob = c(0.2, 0.2, 0.2, 0.1, 0.2, 0.1))makePOIprelim( dateinit = "1975/01/01", scenario = 1, femaleprop = 0.5, ext = 100, numsims = 10000, seed = 123, birthprob = c(0.09, 0.9, 0.01), region = c("North America", "South America", "Africa", "Asia", "Europe", "Oceania"), regionprob = c(0.2, 0.2, 0.2, 0.1, 0.2, 0.1))
Missing person shiny app
mispiApp()mispiApp()
An user interface for computing non-genetic LRs and conditioned probability tables.
CPT_MP()CPT_MP()
This function calculates the Kullback-Leibler divergence for all pairs of provided datasets, considering allele frequencies. It normalizes data, adjusts zero frequencies, and computes KL divergence in both directions for each pair.
multi_kl_divergence(datasets, minFreq = 1e-10)multi_kl_divergence(datasets, minFreq = 1e-10)
datasets |
List of dataframes, each containing allele frequencies for different populations. |
minFreq |
Minimum frequency to be considered for unobserved or poorly observed alleles. |
A matrix containing the Kullback-Leibler divergence for each dataset pair.
kl_matrix <- multi_kl_divergence(list(Argentina, BosniaHerz, Europe))kl_matrix <- multi_kl_divergence(list(Argentina, BosniaHerz, Europe))
postSim: A function for simulating posterior odds
postSim( datasim, Prior = 0.01, PriorModel = c("prelim", "uniform")[1], eps = 0.05, erRs = 0.01, epc = Cmodel(), erRc = Cmodel(), MPc = 1, epa = 0.05, erRa = 0.01, MPa = 10, MPr = 2 )postSim( datasim, Prior = 0.01, PriorModel = c("prelim", "uniform")[1], eps = 0.05, erRs = 0.01, epc = Cmodel(), erRc = Cmodel(), MPc = 1, epa = 0.05, erRa = 0.01, MPa = 10, MPr = 2 )
datasim |
Output from simLRgen function. |
Prior |
Prior probability for H1 |
PriorModel |
Prior odds model: "prelim" is based on preliminary data, and "uniform" uses only the prior probability of H1 |
eps |
epsilon parameter sex |
erRs |
error parameter sex |
epc |
epsilon parameter hair color |
erRc |
error parameter hair color |
MPc |
Missing person hair color |
epa |
epsilon parameter age |
erRa |
error parameter age |
MPa |
Missing person age |
MPr |
Missing person age error range |
A value of posterior odds.
library(forrel) x = linearPed(2) plot(x) x = setMarkers(x, locusAttributes = NorwegianFrequencies[1:5]) x = profileSim(x, N = 1, ids = 2) datasim = simLRgen(x, missing = 5, 10, 123) postSim(datasim)library(forrel) x = linearPed(2) plot(x) x = setMarkers(x, locusAttributes = NorwegianFrequencies[1:5]) x = profileSim(x, N = 1, ids = 2) datasim = simLRgen(x, missing = 5, 10, 123) postSim(datasim)
This function creates a dataframe that lists every unique combination of hair colour, skin colour, and eye colour in the provided dataset, along with the proportion of occurrences of each combination.
refProp(data)refProp(data)
data |
A data.frame containing the characteristics of individuals. |
A data.frame with columns for hair_colour, skin_colour, eye_colour, and f_h_s_y.
data <- simRef(1000) refProp(data)data <- simRef(1000) refProp(data)
simLR2dataframe: A function for extracting LR distributions in a dataframe from simLRgen() output.
simLR2dataframe(datasim)simLR2dataframe(datasim)
datasim |
Input dataframe containing expected LRs for related and unrelated POIs. It should be the output from makeLRsims function. |
A dataframe with LR values obtained from simulations.
Simulate likelihoods ratio (LRs) based on genetic data: a function for obtaining expected LRs under relatedness and unrelatedness kinship hypothesis.
simLRgen(reference, missing, numsims, seed, numCores = 1)simLRgen(reference, missing, numsims, seed, numCores = 1)
reference |
Reference pedigree. It could be an input from read_fam() function or a pedigree built with pedtools. |
missing |
Missing person ID/label indicated in the pedigree. |
numsims |
Number of simulations performed. |
seed |
Select a seed for simulations. If it is defined, results will be reproducible. Suggested, seed = 123 |
numCores |
Enables parallelization |
An object of class data.frame with LRs obtained for both hypothesis, Unrelated where POI is not MP or Related where POI is MP.
library(forrel) x = linearPed(2) plot(x) x = setMarkers(x, locusAttributes = NorwegianFrequencies[1:5]) x = profileSim(x, N = 1, ids = 2) datasim = simLRgen(x, missing = 5, 10, 123)library(forrel) x = linearPed(2) plot(x) x = setMarkers(x, locusAttributes = NorwegianFrequencies[1:5]) x = profileSim(x, N = 1, ids = 2) datasim = simLRgen(x, missing = 5, 10, 123)
Simulate likelihoods ratio (LRs) based on preliminary investigation data: a function for obtaining expected LRs under relatedness and unrelatedness kinship hypothesis.
simLRprelim( vartype, numsims = 1000, seed = 123, int = 5, ErrorRate = 0.05, alphaBdate = c(1, 4, 60, 11, 6, 4, 4), numReg = 6, MP = NULL, database, cuts = c(-120, -30, 30, 120, 240, 360) )simLRprelim( vartype, numsims = 1000, seed = 123, int = 5, ErrorRate = 0.05, alphaBdate = c(1, 4, 60, 11, 6, 4, 4), numReg = 6, MP = NULL, database, cuts = c(-120, -30, 30, 120, 240, 360) )
vartype |
Indicates type of preliminary investigation variable. Options are: sex, region, age, birthDate and height. |
numsims |
Number of simulations performed. |
seed |
Seed for simulations. |
int |
Interval parameter, used for height and age vartypes. It defines the estimation range, for example, if MP age is 55, and int is 10, the estimated age range will be between 45 and 65. |
ErrorRate |
Error rate for sex, region, age and Height LR calculations. |
alphaBdate |
Vector containing alpha parameters for Dirichlet distribution. Usually they are the frequencies of the solved cases in each category. |
numReg |
Number of regions present in the case. |
MP |
Introduce the preliminary data of the selected variable (vartype) of the MP. If it is null, open search is carried out. If it is not NULL, close search LR is computed. Variables values must be named as those presented in makePOIprelim function. |
database |
It is used when the close search (MP not NULL), is carried out. It could be the output from makePOIprelim or a database with the same structure. |
cuts |
Value of differences between DBD and ABD used for category definition. They must be the same as the ones selected for alphaBdate vector. |
An object of class data.frame with LRs obtained for both hypothesis, Unrelated where POI/UHR is not MP or Related where POI/UHR is MP.
library(mispitools) simLRprelim("sex")library(mispitools) simLRprelim("sex")
This function simulates a dataset representing physical characteristics (hair color, skin color, eye color) of a hypothetical population, based on conditional probability distributions. The size of the simulated population can be adjusted by the user.
simRef(n = 1000, seed = 1234)simRef(n = 1000, seed = 1234)
n |
The number of individuals in the simulated population. |
seed |
Selected seed for simulations. |
A data.frame with three columns: hair_colour, skin_colour, and eye_colour,
each representing the respective characteristics of each individual in the sample population.
The hair color is simulated based on predefined probabilities, and skin and eye colors
are generated conditionally based on the hair color.
simRef(1000) # Generates a data frame with 1000 entries based on the defined distributions.simRef(1000) # Generates a data frame with 1000 entries based on the defined distributions.
Threshold rates: a function for computing error rates and Matthews correlation coefficient of a specific LR threshold.
Trates(datasim, threshold)Trates(datasim, threshold)
datasim |
Input dataframe containing expected LRs for related and unrelated POIs. It should be the output from makeLRsims function. |
threshold |
Likelihood ratio threshold selected for error rates calculation. |
Values of false positive and false negative rates and MCC for a specific LR threshold.
library(forrel) x = linearPed(2) x = setMarkers(x, locusAttributes = NorwegianFrequencies[1:5]) x = profileSim(x, N = 1, ids = 2) datasim = simLRgen(x, missing = 5, 10, 123) Trates(datasim, 10)library(forrel) x = linearPed(2) x = setMarkers(x, locusAttributes = NorwegianFrequencies[1:5]) x = profileSim(x, N = 1, ids = 2) datasim = simLRgen(x, missing = 5, 10, 123) Trates(datasim, 10)
STRs allelic frequencies from specified country.
USAUSA
A data frame allele frequencies