Understanding the Paternity Index
Paternity Index and Combined Paternity Index
Each DNA marker that is used in a paternity test is given a Paternity Index (PI) which is a calculated value based upon the inheritance scenario and genotypes that exist. The Paternity Index reflects the weight of the statistical strength of genetic association given the discovered Mother, Child or Alleged Father genotypes.
Broken down, the Paternity Index is simply a ratio of the test of two different hypotheses: Numerator (the probability that the Alleged Father of a certain race is the Child’s true biological father given the genetic evidence seen) vs. Denominator (probability that a random man of a particular race could be the biological father of the child given the genetic evidence seen).
PI must be calculated for each genetic marker (CSF1PO, Penta D, Penta E, FGA, etc.). All PI’s are then multiplied together to get a Combined Paternity Index (CPI). The CPI is then converted to a probability of paternity (such as 99.99%). These are the highest-level of calculation steps that are performed.
There is indeed logic behind calculating a Paternity Index value and this logic that is based upon inheritance.
Let’s step through this scenario:
1. As mentioned above, the PI = Numerator / Denominator. Let’s consider the following example where we have a Mother that has Allele A and Allele B, Child that has Allele A and Allele B and an Alleged Father that has an Allele B and Allele C. This is the same type of example which addresses the question at hand:

2. Let’s break this scenario down to the inheritance and understand how the Numerator is derived:

3. We calculate the Numerator first. Because we know the genotype of the Alleged Father, we perform our Numerator calculations in the context of Mother, Child and Alleged Father’s relationship. We are making the assumption that the Alleged Father could potentially be the biological father given his genotype which means he could have only contributed Allele B to the child. He has a 50% chance of passing Allele B to the Child. If this is true it would also mean that the Mother could only have given Allele A in the CONTEXT of the Child’s and Alleged Father’s genotype. That is the point here. You will see the difference when we calculate the Denominator for an untested, unrelated, random man in the population (Caucasian, Hispanic, etc.)

4. Next, we calculate the Denominator. We do not use the Alleged Father’s genotype here. Again, the Denominator is a hypothesis. The hypothesis asks the question of “What are the chances that a random man could be the biological father given the Mother and Child’s genotypes in conjunction with what we know about the marker and it’s allele frequencies? The Denominator in this example takes into consideration all potential genetic inheritance possibilities given that a random man contains the same genotype as the child….which of course is the perfect scenario because all alleles that are found in the Child are accounted for:

5. In essence, the Numerator captures the following in the equation: [(probability that the Mother passes Allele A (0.5) * the probability that Allele B was passed by a random man) + (probability that the Mother passes Allele B * the probability that Allele A was passed by a random man)] which equates to [Possibility #1 + Possibility #2] above. So you can see that all genetic scenarios are taken into account. Because of this situation, the Paternity Index is affected which subsequently impacts the CPI which in turn changes the Probability of Paternity




