Founder analysis is an attempt to identify the founder responsible for an autosomal recessive disease.  We start by making
the assumption that the defective gene (allele) was introduced by one and only one founder.

There are many different founders in the pedigrees of the GM₁ affecteds.  In order to discriminate
between these founders we can ask:

    Which founders could have contributed an allele to each of the carriers?
    Which founders were the most highly related to the GM₁ group?

We will answer these questions in terms of consanguinity.  The consanguinity between any two dogs is the probability that
those two dogs will share an allele that is identical by descent.  Two dogs that are completely unrelated will have a
consanguinity of 0.  The consanguinity between any two founders is assumed to be 0.  The consanguinity between a founder
and its child is 0.25
 

Table 1 gives an overview of the founders, their relationship to the entire population and their relationship to the population
of GM₁ carriers.  The founders are ordered according to their average consanguinity with the entire population
(second column).   6 founders have fairly significant associations with GM₁

We can look at the structure of the population for the earliest known generations to get a better idea of why so many founders are implicated in the initial analysis.

 The known population structure for the earliest generations of 9 different founders.  The green squares indicate the three founders suspected of introducing the defective GM₁ allele (i.e. Cigana, Lontra, and Dina).  The red squares indicate dogs that are related to another suspected founder, Tabu.

The known population structure for the earliest generations of 9 different founders is shown in figure 1.  The red squares indicate dogs that are related to Tabu.  The green squares indicate the three founders implicated above (i.e. Cigana, Lontra, and Dina).    The three green founders combine to form a common pool which is most easily represented by one dog, Leao.  Leao's line follows a complicated inbred path until the 6th and 7th generation, where it is almost completely mixed with Tabu.  The only descendent of Leao that is not related to Tabu is Lumpi ( in the lower left-hand corner).

One way to distinguish between the genetic legacy of Leao and the genetic legacy of Tabu is through Lumpi.  However, as table 2. indicates there is no evidence that Lumpi was a GM₁ carrier.  Consequently, it is very difficult to discriminate between Leao and Tabu for GM₁.