recombination, shuffling, and revealing recessive alleal varients

[RyGuyFly]

Hi everyone.

 

I'm breeding a small hamster species Phodopus Roborovskii that is relatively speaking new to the pet trade. As a result there are not a lot of genetic morphs or coat colors documented and for the most part they are wild type agouti. My theory that when new phenotype variations pop up it is unlikely that the mutation occurred right then. More likely it occurred and was unknowingly carried for generations until someone crossed two related individuals who happened to be carriers.

 

So I was thinking if I wanted try and reveal some undocumented recessive alleles what would be the best way to go about it. My thought is the best way to test all the alleles would be to breed parent to child. Now due to recombination and reshuffling of the genes in the germ line cells each child would have a slightly different set of the parents genes. So I would need to do parent child mating with several different offspring. Let's say I'm going to try this with father daughter pairings.

 

One of my questions is when the father produces his germ line cells and they are recombined/shuffled do we have any idea what the percentages are? Is it generally 50/50ish? That is to say the father passes on roughly 50% of his genes from is mother and 50% from his father?

 

Say we breed this male to and unrelated female. The genetic make up of his offspring would be possibly roughly 25% from his mother and 25% from his father. But every child would have a different set of these genes that make up those 25%s. As a result I would need to produce several female offspring from this male so that combined they would would have 100% of his genes. Then I could breed those offspring back to the male and roughly 25% of the genes of each offspring would be homogeneous. If I did this with a bunch of the female offspring eventually I would get from the f2 generation examples of homogeneous alleals for every single one of the initial males genes.

 

My question here is how many different female offspring would I need to make sure I randomly covered about 100% of the males genes? My first thought is that I would probably need a ton so that statistically I could be assured of 100 of his genes being passed on. But i have limited space so I was thinking 3 or 4 would probably cover the vast majority but idk. Each female would share many of her genes with her siblings and only a portion would be unique to her. So in the end I'm not sure how many female offspring I need Any thoughts?

 

One of the possible problems I thought of is that if the hamsters are already heavily inbred there may be very little variation to reveal by means of these pairings. I think it would be best to do this experiment starting with a wild male from a genetically diverse population.

 

Thanks.

Edited August 30, 2015 by Ryan1234

[MonDie]

If you can keep them monogamous, then siblings will also be (on average) 50% similar. Father-daughter mating assumes the father carries the mutation, which you do not know. You can't do father-daughter matings anyway unless you know the father.

 

Each cross is an independent event, and the probability of a recessive allele (if it exists) becoming homozygous is one-in-eight for each cross. There is a 0.5 chance the offspring inherited the mutation, in which case it will be a monohybrid cross, which has a 0.25 chance of producing a homozygote. 0.5*0.25 comes to 0.125 per cross. The complementary event of not getting a homozygote thus has a 1-0.125 or 0.875 chance. The odds of not getting a homozygote in two tries is thus 0.875^2; three tries 0.875^3; etc.

 

The longer you stick with one breeding group, the smaller odds you have of finding a recessive mutation per mating. From the above it can be derived that the odds drop by 12.5% each mating. To demonstrate, the difference between 1-.875^2 and 1-.875^3 is only 87.5% of the difference between 1-.875 and 1-.875^2, and even though those numbers assume that there is one and only one recessive mutation present in the breeding group, the rule is universally true.

 

to the title

*allele variants

Edited August 31, 2015 by MonDie

[RyGuyFly]

Hi mondie.

 

Thank you for your response. It is true that all the siblings would inherit 50% of their genes from their father. However the 50% they would inherit from their father would not necessarily be comprised of the same genes. Its my understanding that each sibling is around 25% similar for genes from their father side. Hence the need to use several offspring insuring that 100% of the father's genes are accounted for.

 

You state that the chance of an allele becoming homozygous is one in eight for each cross. I'm not sure I follow. Why isn't it one in four? And then wouldn't my chances of not getting homozygous be .75? Maybe there is something I'm missing?

 

Also remember that each father/daughter mating will produce around 4-6 pups. Each pup being an independent roll of the figurative die. And each of the initial daughter females will share some genes so the figurative dice would be rolled again in other litters for the certain alleles they have in common.

 

However, my question is not what are my chances of getting a homozygous allele. But rather the number of females needed to ensure I get a copy of every (or most every) gene from the father into that initial set of daughter females.

 

Also regardless of the chances for homozygous alleles, if someone wanted to test for hidden recessive alleles I think this would be the most efficient way of going about it. My other thought would be "line breeding" the initial male. That is to say breed him with his daughter, then granddaughter, then great granddaughter etc. Eventually you would get homozygous alleles that way too. The downside is that your getting an increase in homozygous alleles all across the board and the inbreeding coefficient would be high. The way I am currently proposing limits the number of homozygous alleles per individual offspring and makes up for it by using multiple different offspring.

[MonDie]

I assumed one pup per mating. I should have said per offspring instead of per mating.

 

Each parent has 46 chromosomes, adding to 92. Each offspring inherits 46 of these 92, 23 from mother and 23 from father. 46 is half of 92 so each offspring inherits half of the parental gene pool, and doesn't inherit the other half. When comparing the genome of another offspring to its sibling, there is thus 50/50 odds that any particular chromosome will also have been inherited by its sibling, assuming they have the same mother and same father.

 

If either parent carries a recessive mutation, then each offspring has a .50 chance of inheriting it. If it inherits it, it is a hybrid, and crossing it with another hybrid such as the paremt that carried the allele would be a monohybrid cross. CORRECTION: It is actually .25 since there is a .5 chance of being a hybrid for each progenitor, making it .25 chance of a monohybrid cross. This is unavoidable since yoy don't know which parent originally carried the mutation. The monohybrid cross has a .25 chance of producing a recessive homozygote. The chance of both events happening is thus .5 .25 times .25, which comes to .125 .0625.

 

When you mate an offspring to parent, the resultant offspring is 75% similar to each parent. That means .75*.0.625 when you mate the daughter/grandchild to its father/grandfather. Inbreeding isn't necessarily desireable however since it will result in more spontaneous abortions... plus ethics.

Edited August 31, 2015 by MonDie

[RyGuyFly]

Thanks for clarifying. I understand what you are saying now. So basically there 6.25% chance that any given allele will be passed to the daughter and then become homozygous in the grand daughter. That's a relatively low % but we are talking about a multitude of pups, an unknown quantity of recessive alleles in the father and unknown effects when the alleles are homozygous.

 

I am attempting a tri-hybrid cross were both parents are heterozygous for 3 different alleles. The chance of getting a single offspring that is homozygous for all 3 genes is only .0156. So .0625 is better chances than .0156 lol

 

Anyways thanks for your help.

[MonDie]

My final paragraph got completely minced during correcting. It's actually .75*.5*.25 = .09375

.5 that the parent carrying the allele is the one selected for breeding.

 

Perhaps you can counteract the inbreeding with health promoting techniques. For example try to breed the healthiest picks of the litter. Be equally watchful for potentially recessive traits that are harmful to health, and avoid inbreeding those alleles.

Edited September 1, 2015 by MonDie

[RyGuyFly]

Yea exactly. I feel my little experiment is definitally doable in an ethical way and i plan to try it out. Over the next year I will have produced many F1 and f2 pups to choose from. The hopes are that this will produce some amazing new coat color or texture. Its a long shot tho and pretty unlikely. I think more likely I might reveal some more minute change like odd shaped ears or tail. Or some small change in body shape. Worst case scenario my little experiment reveals nothing or only reveals some change that is completely undesirable.

 

In that case I will move on to my 2nd little experiment which is to breed for what they call continuous or quantitative traits. That's a subject for a whole different thread tho. Lol probably a thread that's shortly forthcoming because I have questions about how to quantify various continuous or quantitative traits. In theory it is doable but being able to examine each animal for a certain trait and give a number value to said trait will be a challenge.

 

I think in labs they will mix sperm cells with carcinogenic materials or expose them to radiation to speed up the processes of creating mutations. This is something I've also thought about. Lol but it might be a challenge to carry out at home. So for right now I'm just banking on natural variations/mutations for my first experiment.