Paternity Calculations
In paternity calculations the likelihood ratio is the same as the paternity index. The three tables below show the calculations of the likelihood ratio for different possible combinations of genotypes assuming that the mother, alleged father and father are unrelated.
In the tables
- GC is the child’s genotype;
- GM is the mother’s genotype;
- GAF is the alleged father’s genotype;
- Hp is the mother and alleged father are the true parents;
- Hd is the mother and an unrelated male are the true parents;
- LR numerator is P(GC|GM, GAF, Hp);
- LR denominator is P(GC|GM, GAF, Hd).01
Likelihood Ratio and Paternity Calculation |
GC |
GM |
GAF |
Num |
Denominator |
LR |
AiAi |
AiAi |
AiAi |
1 |
pi |
1/ pi |
|
|
AiAj, j≠ i |
1/2 |
pi |
1/2pi |
|
|
AjAk, k≠i, j |
0 |
pi |
0 |
AiAi |
AiAj, i ≠ j |
AiAi |
1/2 |
pi /2 |
1/ pi |
|
|
AiAj, j≠i |
1/4 |
pi /2 |
1/2 pi |
|
|
AjAk, k≠i, j |
0 |
pi /2 |
0 |
AiAj, i≠j |
AiAi |
AjAj |
1 |
pj |
1/pj |
|
|
AjAk, k≠j |
1/2 |
pi |
1/2 pj |
|
|
AkAl, k, l≠j |
0 |
pi |
0 |
AiAj, i≠j |
AiAj, i≠j |
AiAi |
1/2 |
(pi + pj)/2 |
1/(pi + pj) |
|
|
AiAj |
1/2 |
(pi + pj)/2 |
1/( pi + pj) |
|
|
AjAk, k≠j |
1/4 |
(pi+ pj)/2 |
1/(2(pi + pj)) |
|
|
AkAl, k, l≠i, j |
0 |
(pi + pj)/2 |
0 |
AiAj, i≠j |
AiAk, k≠i, j |
AjAj |
1/2 |
pj/2 |
1/ p j |
|
|
AjAl, l≠j |
1/4 |
pj/2 |
1/2 pj |
|
|
AkAl, k, l≠j |
0 |
pj /2 |
0 |
Example:
Locus |
Mother |
Child |
Alleged Father |
D3S1358 |
16 |
16 |
15, 16 |
vWA |
17, 18 |
17, 18 |
18, 20 |
FGA |
25, 27 |
23, 25 |
21, 23 |
D18S51 |
14, 18 |
18 |
18, 20 |
The following frequencies were used:
Locus |
Allele |
Frequency |
Allele |
Frequency |
D3S1358 |
16 |
0.2315 |
|
|
vWA |
17 |
0.2627 |
18 |
0.2219 |
FGA |
23 |
0.1581 |
|
|
D18S51 |
18 |
0.0918 |
|
|
- D3S1358 — LR = 1/2p16 = 1/(2 x 0.2315) = 2.1598
- vWA — LR = 1/[2(p17 + p18)] = 1/[2(0.2627 + 0.2219)] = 1.0317
- FGA — LR = 1/2p23 = 1/(2 x 0.1581) = 3.1625
- D18S51 — LR = 1/2p18 = 1/(2 x 0.0918) = 5.4466
Therefore, LR = 2.1598 x 1.0317 x 3.1625 x 5.4466 = 38.3815
In other words, it is 38 times more likely that the child’s DNA profile would be observed if the alleged father is the true father rather than an unrelated individual selected at random from the Caucasian population.
Click here to watch a video on paternity calculations presented by Greggory LaBerge.
< Previous Page :: Next Page >