INTERPRETATION AND REPORTING

• Jurisdiction: Australia

Introduction ................................................................................................... [82.800]

Declaring an exclusion .................................................................................. [82.820]

Population genetic models

Introduction ................................................................................................... [82.900]

Hardy-Weinberg equilibrium.......................................................................... [82.920]

Linkage equilibrium ....................................................................................... [82.940]

Choice of population genetic models............................................................ [82.960]

Population frequencies ................................................................................. [82.980]

Random Man Not Excluded (RMNE)............................................................ [82.1000]

Paternity index .............................................................................................. [82.1020]

Relative chance of paternity ......................................................................... [82.1040]

Interpreting Mendelian inconsistencies ......................................................... [82.1060]

Deficiency cases ........................................................................................... [82.1080]

Disaster Victim Identification ......................................................................... [82.1100]

Relatives defence ......................................................................................... [82.1120]

Mixed foetal material ..................................................................................... [82.1140]

Linked loci ..................................................................................................... [82.1160]

Reporting standards ...................................................................................... [82.1180]

[82.800] Introduction

The interpretation and reporting of parentage or other familial testing type work involves a number of steps and key decisions. These will be outlined in this section.

The first step is typically one of declaring an exclusion or not. We discuss below the concept of an exclusion, its practical utility and the fact that strictly an exclusion cannot be defined with complete accuracy. Many laboratories use the term "declared an exclusion" which has a more robust interpretation, since this term reflects the nature of the practical decision to describe the case in this way. Lawyers reading this paragraph may be amazed that there exists ambiguity about whether or not a case is indeed an exclusion. The ambiguity is often very minor but it is worthwhile being explicit about where this ambiguity exists.

If the case has not been declared an exclusion, then the next step is typically to proceed to a statistic to express the weight of the evidence. Three statistics are in usage and each of these three may employ one of two population genetic models in their generation. There is considerable consensus about the pros and cons of each statistic and about the performance of the population genetic models. However, diversity of approach still exists. The key issues then are which statistic or statistics to report and which population genetic model to utilise. It suits the flow of the later sections to deal with the population genetic models first and the various statistics later. Also of crucial importance is the possibility of a relative as the true father.

Having produced a statistic there are key issues about effective phrasing of the report and especially the meaning of the statistic.

[82.820] Declaring an exclusion

In cases of paternity, it is typically assumed that the mother is the biological mother and, therefore, half of the alleles that have been inherited by the child must be from her.

TABLE 3 Obligatory alleles in a hypothetical case

Locus	D3S1358	vWA	D8S1179	D21S11	D18S51	D5S818	D13S317	D7S820
Father	15-15	15-15	13-13	30-31	16-18	12-12	11-11	9-10
Child	15-17	15-17	11-13	30-31	15-16	12-12	8-11	10-10
Mother	15-17	16-17	11-12	28-30	13-15	11-12	8-12	10-11
Obligate paternal allele	15 or 17	15	13	31	16	12	11	10

Looking at Table 3, in the vWA locus it can be seen that the child must have inherited the 17 allele from the mother (the mother can only pass on a vWA 16 or 17 allele). Therefore, the vWA 15 allele must have been inherited from the biological father and this is known as the obligate or paternal allele. We would therefore expect the true father to possess each of the obligate alleles. This principle is used to draw up a list of obligate alleles derived from our typing results for each case. This list can then be compared with the typings obtained from the putative father.

Whatever typing systems are used, the end result is the same: a series of obligate alleles which the putative father either does or does not carry. If the alleged father does not carry the required allele at a locus, this is described as a Mendelian inconsistency.

There are two types of Mendelian inconsistency (MI):

(1) (1) First-order MI: Figure 6 shows an MI at the vWA locus because the 14 allele must be inherited paternally, but the alleged father has only a 16 or 17 allele (the child has inherited the 18 allele from the mother).

Figure 6 - First-order Mendelian inconsistency

vWA 16-17 vWA 15-18 vWA 14-18

(2) (2) Second-order MI: Figure 7 shows a second order MI at the D21S11 locus. It would appear that the child must have inherited a D21S11 30 allele from both parents (assumed genotype: 30-30). On this basis, an MI exists since the alleged father apparently has the genotype 28-28 (and is expected to pass a 28 allele on to all of his offspring).

Figure 7 - Second-order Mendelian inconsistency

D21S11 (28-28 or 28-X) D21SS11 30-30 D21S11 (30-30 or 30-x)

A first-order MI may result from two possible causes. First, the alleged father may not be the true father or second, he may be the true father but a mutation has occurred. A second-order MI may result from these two causes or from a third. This third explanation results from the presence of a low frequency rare or silent allele.

In the example above, the father may carry an allele which is not amplified during the PCR process because the primer binding site has been altered through a point mutation. Thus the father does not possess two copies of the 28 allele as assumed, but rather carries a 28 allele and a mutated allele (x in Figure 7). It is the mutated allele which he has passed on to the child, whose genotype therefore is 30-x and not 30-30 as assumed. As the mutation has occurred at the primer binding site the allele has not been amplified and is not detected (see the section on "Technology used to perform DNA typing" at [82.600]).

The term null allele has been used extensively in the past but the term silent is preferred as the allele is present but cannot be visualised.

Clearly an alleged father who possesses all the obligate alleles is not excluded. However, males who show one or a few MIs are also not excluded, although the genotype of the child is less likely if they are the father. Taken to the absolute extreme, if we allowed any number of MIs then no male is excluded. This is the ambiguity surrounding the declaration of an exclusion. The probability of the child's genotype becomes very small if there are a few MIs and for practical purposes, most laboratories declare an exclusion at a pre-determined number of MIs. Those laboratories using a threshold approach to the number of Mendelian inconsistencies will declare an exclusion if the number of inconsistencies exceeds their threshold. This number may be one, two or three MIs and should depend on the total number of loci examined, the mutability of the loci and the frequency of silent alleles in the populations of interest. Clearly, one MI out of a panel of six loci should be viewed differently from one MI out of a panel of 10,000 loci.

Strictly any male not excluded is described as an "inclusion".

Hallenberg and Morling (2002) report the requirements from a number of laboratories for issuing a report with a positive weight for paternity (Table 4). This shows the percentage of laboratories using a predetermined threshold for the allowable number of Mendelian inconsistencies. As can be seen, not all laboratories report such a requirement. They may be using such thresholds in practice or as is discussed later in this section (see "Interpreting Mendelian inconsistencies" at [82.1000]), it is possible to avoid the use of a threshold at all.

TABLE 4. Requirements for issuing a report with a positive weight for paternity

Required paternity index	Required probability of paternity	2000 (N=33)% of labs	2001 (N=36)% of labs
100-1000	99-99.9%	24	19
1000-10,000	99.9-99.99%	33	22
10,000-100,000	99.99-99.999%	21	25
>100,000	>99.999%	3	8
Fewer than a certain number of Mendelian inconsistencies		9	11
No requirement		9	14

As shown in Table 4 above, requirements for issuing a report with a positive weight for paternity varies. For example some laboratories require a PI...