Risk Assessment - by Jim Seltzer

Subject: Risk Assessment - part 1

George A. Padgett, DVM in his book, "Control of Canine Genetic Diseases," Howell Book House, 1998, discusses the use of pedigrees to evaluate the risk that a dog is affected or a carrier. Padgett develops algorithms to trace the transmission of risk from one generation to the next. However, he makes several simplifications that can introduce errors. These posts are an attempt to provide corrections. The math is tedious though not difficult.

Start by laying out the pedigree for the proposed breeding. As an example consider:

```
Old Joe
Harry
Jacqueline
(Puppy)
Old Sam
Tabatha
Jacqueline

```

which is a half-sib mating.

Next, mark the attributes next to the names as follows:

A affected
C carrier (but not affected)
P has produced an affected
U not affected (might be a carrier)
N clear

```
Old Joe (P)
Harry
Jacqueline (C)
(Puppy)
Old Sam
Tabatha (U)
Jacqueline(C)

```

We have no information about Old Sam or Harry, and would like to calculate the probabilities that the puppy will be (1) affected, or (2) a carrier.

We shall assume that the defect is a simple autosomal recessive, and that the average number of affecteds in the population pool from which all the grandparents are drawn is 1%. Assuming Hardy-Weinberg equilibrium, we observe that:

p = defective gene frequency = Square Root(rate of affecteds)
= SQRT(.01) = .1
q = normal gene frequency = 1 - p = .9
2 x p x q = carrier rate = 2 x .1 x .9 = .18 = 18%
q x q = clear rate = .9 x .9 = .81 = 81%

Note that the population background:

affected rate + carrier rate + clear rate =

.01 + .18 + .81 = 1 (or 100%) as it must.

***************************
Jim Seltzer
Willowind Dalmatians
http://users.nbn.net/~jseltzer
****************************