Reptile & Amphibian Forums

Hi all,

I know het status can be proven *IF* you observe fully recessive offspring from a cross with another known recessive or het morph. The flip side is that you can't absolutely prove homogenous genes though probablistic means should give you a pretty good guess if the sample is large enough.

Question is... how hard would you try to confirm het status before giving up? Say you have a 50% chance of getting albino offspring from a het x albino crossing. How many "normal looking" offspring would you go through before declaring the original "potential het" to be a non-het?

Granted, you could always wait for the normal-looking offspring from the "potential het" x albino to grow up (giving you your hets) but that puts you a generation behind schedule and may not work if you're only breeding to the albino to prove or disproving het status of the parent for that particular gene.

N.F.

>Question is... how hard would you try to confirm het status before giving up? Say you have a 50% chance of getting albino offspring from a het x albino crossing. How many "normal looking" offspring would you go through before declaring the original "potential het" to be a non-het?

I would keep trying until the probability is less than 1% that it's a heterozygote that hasn't produced any albinos simply through the luck of the draw. The way to determine that is figure out the probability of getting one normal, which in this case is 50% or 0.5. The probability of getting all normals is 0.5 to the nth power, where n is the number of babies.

Probability of getting all normals
1 baby = 0.5
2 babies = 0.5 x 0.5 = 0.25
3 babies = 0.5 x 0.5 x 0.5 = 0.125
4 babies = 0.5 x 0.5 x 0.5 x 0.5 = 0.0625
5 babies = 0.5 x 0.5 x 0.5 x 0.5 x 0.5 = 0.03125
6 babies = 0.5 x 0.5 x 0.5 x 0.5 x 0.5 x 0.5 = 0.015625
7 babies = 0.5 x 0.5 x 0.5 x 0.5 x 0.5 x 0.5 x 0.5 = 0.0078125

So in this mating, if I got seven or more babies, and all of them were normal, then I would class the possible het as a normal and move on to a new project.

The same method is used when breeding a possible het to a known het. Only the probability of getting a normal from this mating is 0.75. And the probability of getting all normals is 0.75 to the nth power. A calculator that can do powers is very handy for this sort of thing.

Paul Hollander

Thanks for the info. It's not quite as horribly difficult as I once supposed (at least for the example above). Being able to prove that a gecko was double het sounds possible without having too many hatchlings, especially if you test by breeding to a pure blizzard and to a pure patternless. Once proven, the double het could be used in other projects earlier than it's offspring can grow up.

Cheers!

N.F.