BTB, Comparison Reports, Statistics

Question 1

What PAS does vs DPAA

Answer 1

PAS: performs requested DNA testing and submit comparison resquest; use match criteria to determine results of mtDNA comparisons; writes BTB reports NOT identification reports
DPAA: sends requests for DNA testing, does official identification from supported DNA, sends comparison request

Question 2

To complete LISA comparison

Answer 2

• LISA>Case Management> mtDNA g:cat Search/Stats > B-P(batch to pairwise)
• selected “comp reg?”
• select case
• search All or Asg

Question 3

To complete BTB

Answer 3

• gather all files
• print or open CoC
• rerun ALL comparisons
• run and print CPD g:cats for each sample and maternal FRS
• run and print FRS database g:cat if Blind Hit(not for refs)
• print summary and detail
• print FRS Servicemember Relationship Report; cross out invalid or duplicate results
• DON’T do Staff g:cat
• complete and print LSAM - likelihood ratio stats (reverse of the profile probability - report 3 sig figs)

Question 4

LSAM

Answer 4

• Clopper Pearson calculator that calculates profile probability for data (set theta to 0)
• uses counting method to count number of times sequence is observed in population
• apply 95% confidence interval
• LR truncated to whole number

Question 5

Theta

Answer 5

• set to 0 for mtDNA
• accoutns of p^2 in HWE to correct for less genetic variability
• gives indication of population substructure
• we can reasonably assume in a substructure the allele frequencies would be different than we typically account for

Question 6

Likelihood ratio

Answer 6

-LR
-inverse of the probability
-when two mtDNA sequences from separate sourches match, the two sources cannot be excluded as being from the same person
-LR = Pr(E|H1)/
Pr(E|H2)
-Probability of the profile, or evidence, given the DNA is that of the service member over the probability of the profile, or evidence, given the DNA matches a random person

Question 7

Hardy-Weinberg Laws

Answer 7

• in large, random-mating population, with no disturbances like mutation, migration, or selection, the relative proportions of different genotypes remain constant generation to generation (is for random match probability)
• 2 allele system: p2+2pq+q2=1

Question 8

Kinship Index

Answer 8

familial relationships; allele frequencies showing the frequency of shared allele (mother & son
-LR per locus multiplied together

Question 9

Counting method

Answer 9

• basis for mtDNA and Y-STR haplotype frequency estimation
• don’t just report b/c: CPD search is counting method and gives haplotype frequency for mtDNA and Y-STR; apply 95% confidence interval to be more conservative
• input CPD#(caucasion, African American, Hispanic database) and put into clopper pearson in LSAM
• confidence interval accounts for database size and sample variation

Question 10

Product Rule

Answer 10

• combination of locus-specific match probabilities to get a profile frequency estimation
• match probability for one locus combined with additional loci to decrease the odds of a random match to an unrelated individual (when there is linkage equilibrium)
• when alleles at different loci are independent of each other

Question 11

Blind hit vs Non-blind hit

Answer 11

• blind: CPD g:cat for references and samples; CPD FRS for samples (FRS servicemember relationship report; blind hit template
• non-blind: CPD g:cat for references and samples; no g:cat FRS fro samples; non-blind hit template

Question 12

Clopper Pearson

Answer 12

-95% confidence interval used to account fo rlimited population size b/c database doesn’t hav eall sequences
-LSAM>statistical analysis>clopper pearson
-1-tailed upper
-want profile probability and take inversion to get likelihood ratio
-use 3 sig figs with statistical statement
“likelihood of data is X times more likely to be related to the ref names than being an unrelated individual in (Caucasion, AA, Hispanic) population group.”

Question 13

g:cat example
#1: 16024-16391/35-369
16093C, 16193.1C, 16311C, 73G, 263G, 315.1C
#2: 16024-16391/35-369
16311C, 73G, 263G, 309.1C, 315.1C

Answer 13

g:cat results: difference at 16093C so wouldn’t show on g:cat

Question 14

Comparison request

Answer 14

• compares mito(maternal) references to all samples in H case (NOT a BTB)
• run in eLISA or manually (g:cat function w/ B-P)
• VR2 and PS5 tested separate from HV1 and 2 as an addendum
• uses match criteria to determine results to compare overlapping regions
• if 16093 is a difference, must have 2 other differences to exclude

Question 15

What will eLISA tell about BTB report?

Answer 15

• if it’s a priority
• blind hit or non-blind hit report
• if any remaining sample in house to be “returned” to DPAA or “consumed” for training/validation
• name of SVM
• relevant case and evidence