Lecture 15 Individual Identity and Paternity Analysis October

Lecture 15: Individual Identity and Paternity Analysis October 15, 2012

Lab 7 Update u Corrected instructions for lab 7 are on lab webpage u Up to 3 points extra credit if problem 2 is done correctly u New version of problem 2 can be answered with analysis that most people completed in lab u See lab open hours schedule on lab web page u http: //www. as. wvu. edu/~sdifazio/popgen/lab. html

Last Time § Interpretation of F-statistics § More on the Structure program § Principal Components Analysis § Population assignment

Today § Population assignment examples § Forensic evidence and individual identity § Introduction to paternity analysis

Population Assignment: Likelihood § "Assignment Tests" based on allele frequencies in source populations and genetic composition of individuals for homozygote Ai. Ai in population l at locus k § Individual likelihood often summarized as: -log 10(P(G|Hn)) for heterozygote Ai. Aj in population l at locus k § Low numbers mean higher probability of Hn § Likelihoods can be plotted against each other for m loci

Population Assignment Example: Wolf Populations in Northwest Territories § Wolf populations sampled on island mainland populations in Canadian Northwest Territories -log likelihood from Mainland § Immigrants detected on mainland (black circles) from Banks Island (white circles) -log likelihood from Banks Island Carmichael et al. 2001 Mol Ecol 10: 2787

Population Assignment Example: Fish Stories § Fishing competition on Lake Saimaa in Southeast Finland § Contestant allegedly caught a 5. 5 kg salmon, much larger than usual for the lake § Officials compared fish from the lake to fish from local markets (originating from Norway and Baltic sea) § 7 microsatellites Lake Saimaa Market § Based on likelihood analysis, fish was purchased rather than caught in lake -

Individual Identity: Likelihood § Assume you find skin cells and blood under fingernails of a murder victim § A hitman for the Sicilian mob is seen exiting the apartment § You gather DNA evidence from the skin cells and from the suspect § What is H 1 and what is H 2?

Likelihood ratio forensic DNA evidence § Matching a suspect to a piece of physical DNA evidence requires comparing likelihoods of two alternative hypotheses: Ø H 1 is that the sample comes from the suspect Ø H 2 is that the sample is from a person unrelated to the suspect Ø E is the DNA evidence § P(E|H 1) =1 if the DNA profile of the suspect matches the DNA profile of the evidence § P(E|H 2) is probability of observing the suspect genotype in an unrelated person in the population (i. e. , the match probability)

Genetic typing in forensics § Highly polymorphic loci provide unique ‘fingerprint’ for each individual § Tie suspects to blood stains, semen, skin cells, hair § Revolutionized criminal justice in last 20 years § Also used in disasters and forensic anthropology § Principles of population genetics must be applied in calculating and interpreting probability of identity

Markers in Genetic Typing § Many sets of markers commonly used § Sets of highly polymorphic microsatellites (also called VNTR (Variable Number of Tandem Repeats), STR (Short Tandem Repeat) or SSR (Simple Sequence Repeat)) : http: //www. cstl. nist. gov/div 831/strbase//mlt_abiid. htm Ø Profiler Plus, SGM Plus, Cofiler, Identifiler, Power. Plex 16 § Most are amplified in a single multiplex reaction and analyzed in a single capillary § Very high “exclusion power” (ability to differentiate individuals)

Match Probability § Probability of observing a genotype at locus k by chance in population is a function of allele frequencies: Homozygote Heterozygote for m loci § Assumes unlinked (independent loci) and Hardy. Weinberg equilibrium

Probability of Identity § Probability 2 randomly selected individuals have same profile at locus k: Homozygotes Heterozygotes for m loci § Exclusion Probability (E): E=1 -P

What if the slimy mob defense attorney argues that the most likely perpetrator is the mob hitman’s brother, who has conveniently “disappeared”? Does the general match probability apply to near relatives?

Probability of identity for full sibs Homozygotes 0 alleles IBD 2 alleles IBD 1 allele IBD Heterozygotes 2 alleles IBD 0 alleles IBD General Probability of Identity for Full Sibs:

Probability of identity for full sibs Probability of identity unrelated individuals For a locus with 5 alleles, each at a frequency of 0. 2: PID = 0. 072 PIDsib = 0. 368

What is minimum probability of identity for full sibs?

Application to bear and wolf populations § Populations have high relatedness within restricted areas: family structure Probability of Identity § Collect anonymous hair samples in wild populations Loci Woods et al. 1999

Which allele frequency to use? § Human populations show some level of substructuring ØFST generally < 0. 03 § Challenge is to choose proper ethnic group and account for gene flow from other groups Illinois Caucasian Georgia Caucasian U. S. Black http: //books. nap. edu/openbook/0309053951/gifmid/95. gif

Substructure in human populations § GST is quite high among the 5 major groups of human populations for CODIS microsatellites § Relatively low within groups, but not 0!

NRC (1996) recommendations § Use population that provides highest probability of observing the genotype (unless other information is known) § Correct homozygous genotypes for substructure within selected population (e. g. , Native Americans, hispanics, African Americans, caucasians, Asian Americans) § No correction for heterozygotes Homozygotes Heterozygotes

Why is it ‘conservative’ (from the standpoint of proving a match) to ignore substructure for heterozygotes?

Example: World Trade Center Victims § Match victims using DNA collected from toothbrushes, hair brushes, or relatives § Exact matches not guaranteed Ø Allele dropout § Use likelihood to match samples to victims

Parentage Analysis: Paternity Exclusion § Determine multilocus genotypes of all mothers, offspring, and potential fathers § Determine paternal gamete by “subtracting” maternal genotype from that of each offspring. § Infer paternity by comparing the multilocus genotype of all gametes to those of all potential males in the population § Assign paternity if all potential males, except one, can be excluded on the basis of genetic incompatibility with the observed pollen gamete genotype § Unsampled males must be considered

Paternity Exclusion § First step is to determine paternal contribution based on seedling alleles that do not match mother Ø Notice for locus 3 both alleles match mother, so there are two potential paternal contributions § Male 3 is the putative father because he is the only one that matches paternal contributions at all loci Locus 1 NO NO YES YES Locus 2 NO NO Locus 3 YES NO NO

Paternity Exclusion Analysis Possible outcomes: Consequences: § Only one male cannot be excluded Female Male § More than one male cannot be excluded Female Male § Analyze more loci Male § All males are excluded Female Male ? § Paternity is assigned Male § Conclude there is migration from external sources

Probabilities of Paternity Exclusion, Single Locus, 2 alleles, codominant § The paternity exclusion probability is the sum of the probability of all exclusionary combinations Hedrick 2005 Probability of a falsely accused male of not matching for at least one of m loci: See Chakraborty et al. 1988 Genetics 118: 527 for a more general calculation of exclusion power

Alleles versus Loci § For a given number of alleles: one locus with many alleles provides more exclusion power than many loci with few alleles Ø 10 loci, 2 alleles, Pr = 0. 875 Ø 1 locus, 20 alleles, Pr=0. 898 § Uniform allele frequencies provide more power

Characteristics of an ideal genetic marker for paternity analysis § Highly polymorphic, (i. e. with many alleles) § Codominant § Reliable § Low cost § Easy to use for genotyping large numbers of individuals § Mendelian or paternal inheritance