Genetics calculator variation - a request......

[Rich Z]

I know there are several genetics calculator programs on line that do a great job of telling you what sort of results you can expect when you breed sample A with sample B, but more and more I am finding that I need another variant of this calculator. What I would like, and I think would be helpful to others, is a program that will be kind of a reverse genetics calculator. In other words, say someone claims that they have a Hypo Lavender Motley het Sunkissed Blood Red. What are the possibilities of the PARENTS needed to produce something like this?

Yeah, I could sit down and think about it to come up with answers, but I'm getting lazier and want a TOOL that can do this for me.

So, has anyone already done something like this and I just haven't stumbled upon it?

Thanks!

 

[Susan]

You've got fantastic combos in secret tubs hidden in your closet and have completely forgotten how you made them, haven't you?!

 

[HanneysCorn]

I can certainly see where you're coming from here, Rich. I think the concept would be fabulous but I just don't know if it would/could be feasible.

Like in your example...hypo, lavender and mot are homogenously found in the offspring. The parents could well be homo too for each gene in the phenotype, as we know, but they could also carry two, or even just one of the genes from the homogenous parents and the rest could just be lucky recessives matching up. Then you've got those hets the hatchling now has inherited. Were they again from a homo parent of that gene, ie sunkissed, blood red or a lucky recessive catch again.

In order for the calculator to really give an answer it would have to give all the many possibilities.

Susan, I just can't see Rich not knowing every detail of his secret projects; he's been doing this too long...

Great food for thought...having the 'hindsight' already mapped out would make the goal a lot less fuzzy as you begin. Would also give one plenty of ideas for matchups to acheive a desired goal much more quickly by increasing the possibilities.

Something to definitely think about if one is both genetically inclined and knowledgeable about creating such a program.

Ruth

 

[Rich Z]

You've got fantastic combos in secret tubs hidden in your closet and have completely forgotten how you made them, haven't you?!

Actually no......

Ever now and again, I will see or hear of someone advertising something for sale, that when I think of what is necessary to have in order to make such a thing, just seems highly unlikely. Or someone will ask me for something they want and I have to try to explain what it would take to create such a thing, and why it just is not possible for me at this time, not improbable for someone else based on likely combos people are working with. Yes, I do get people asking me for the impossible (at this time) quite often. Possible hets are easy to come by. But the GUARANTEED items are a different story entirely. Soooo, I think a program for people to plug in what they want and then see what is needed to produce such a thing may have a really interesting tool to not only give them a road map of how to get there, but also give them a good idea of how likely it is that such a thing would be available and for sale right now, much less even probable of existing in the first place. As an aside, if someone does claim to have such a thing, then their knowing what adults this person would have to have in order to produce them would certainly go a long way to establishing their credibility. Yes, folks, I know this will come as a shock, but there are people out there who WILL lie about the heterozygous genes an animal is supposed to be carrying. Quite frankly, the ONLY way you can prove that it is not what they say it is, is by determining what adults they have that have supposedly produced it. Even if you breed those animals and do not get what you are supposed to as a result, that STILL is not proof positive that you got snookered. But who the heck wants to waste all that time and effort when knowing beforehand the likelihood of that genetic mix may save someone one heck of a lot of headaches.

And yeah, I know it would be a challenging program to write. But certainly not impossible to do.

 

[tyflier]

Well...it could work. If the answers were given as percentages of possiblities, much like the current calculators out there are, you could at least have a reasonbale "guesstimate".

And if you, like Susan suggested, "forgot" something, you could easily figure out the which possibility it was based on the snakes you have available...

As well...someone just getting started could type in the final outcome they desire, and the calculator could easily spit out the most likely scenarios required to achieve that goal.

 

[Rich Z]

As well...someone just getting started could type in the final outcome they desire, and the calculator could easily spit out the most likely scenarios required to achieve that goal.

Yeah, it may open up some people's eyes as to the BEST way to accomplish a goal. For instance, Breeding Hypo Lavender to Hypo Blood Red is a WHOLE LOT better than breeding Hypo Lavender to Blood Red, or Lavender to Hypo Blood Red, in order to eventually get Hypo Lavender Blood Reds. But that's pretty simple to see. But suppose someone were interested in finally producing a Hypo Lavender Blood Red Caramel instead? What's the BEST road to take to reach that goal? :grin01:

 

[Menhir]

Are you sure that you are aware of all the possibilities?
Just mathematically... I'm quite sure that I could programm such a thing without any problems, but I am not sure if you really want to see the result.

 

[El Jefe]

As an aside, if someone does claim to have such a thing, then their knowing what adults this person would have to have in order to produce them would certainly go a long way to establishing their credibility. Yes, folks, I know this will come as a shock, but there are people out there who WILL lie about the heterozygous genes an animal is supposed to be carrying. Quite frankly, the ONLY way you can prove that it is not what they say it is, is by determining what adults they have that have supposedly produced it. Even if you breed those animals and do not get what you are supposed to as a result, that STILL is not proof positive that you got snookered. But who the heck wants to waste all that time and effort when knowing beforehand the likelihood of that genetic mix may save someone one heck of a lot of headaches.

And hence the valuable tool the ACR can be. Sure people can lie about what adults they are using and registering, but once it is posted it is there and people can see it...no changes a few years later to make offspring something they are not.

I always find it funny how many people out there (particularly in the boa and python world) have all these possible hets that never seem to turn out het. Makes me wonder if they ever really had that possibility of being possible hets? Kind of like every pretty normal corn is an okeetee corn....:sidestep:

 

[pewter]

But suppose someone were interested in finally producing a Hypo Lavender Blood Red Caramel instead? What's the BEST road to take to reach that goal? :grin01:

I would probably use lavenderbloodred x amberbloodred :santa:

 

[Rich Z]

I would probably use lavenderbloodred x amberbloodred :santa:

But wouldn't Hypo Lavender Blood Red x Amber Blood Red be even better? :grin01: That way you only have double het recessives instead of triple het recessives.

 

[pewter]

I know that :crazy02:

 

[paulh]

Corn snake mutants are all recessives, as far as I know. To get any given combination of recessive mutants, both parents must have at least one of each mutant gene the babies must be homozygous for. Okeetee, blood red, and Miami phase plus recessive mutant combos are more difficult.

Seems to me that the most useful reverse calculator would take two inputs: a starting with these snakes imput and an ending with those snakes input. And from that, figure the most efficient plan to get the desired result. Programming such a beast is far beyond me, though I can do it with paper and pencil.

 

[Menhir]

Simple Example:

Anery

- Anery X Anery
- Normal het. Anery X Anery
- Normal het. Anery X Normal het. Anery

(Just to be sure, that there is NO known het - otherwise we could go on using Ghost, Snow, etc. etc.)

Okay, let's get one step further:

Snow
- Snow X Snow
- Snow X Anery het. Amel
- Snow X Amel het. Anery
- Snow X Normal het. Anery, Amel
- Anery het. Amel X Anery het. Amel
- Anery het. Amel X Amel het. Anery
- Anery het. Amel X Normal het. Anery, Amel
- Amel het. Anery X Anery het. Amel
- Amel het. Anery X Amel het. Anery
- Amel het. Anery X Normal het. Anery, Amel
- Normal het. Anery, Amel X Normal het. Anery, Amel

...now imagine a few hets or triple recessive morphs. Thats not a monster, the program is easy, but the output of the program is so huge, that it seems useless to have that list.

The most efficient plan is not available, cause the result is a set of efficient plans. I will explain why - the most efficient pairing is desired morph X desired morph, right? The least expensive pairing would be e.g. Normal het. Anery, Amel X Normal het. Anery, Amel or the P generation only consisting of all genes that are needed. (one could bring this down to a set cover problem, but I think I'm drifting away....)
Okay, so, what we have is a so called multi-objective optimization, cause with having objectives like the ratio of the desired morph in the F2, the cost for the parents, the risk (imagine poss.hets.) and so on, there is no such thing like a single efficient or optimal solution. The goal to find is a so called Pareto-optimal set.
So, the user should get a set of results. Okay?
Lets go one step further: To evaluate each combination, we need to model costs for each morph and also if the breeder already has them. So, we would need a specification of animals including their sexes.

...hmmm, I could bring the whole thing down to a program and maybe I will do that as soon as I have time. I don't think that anyone will use it, but my research topic is multi-objective optimization using evolutionary algorithms, and so I'm quite familiar with that kind of problems. I just want to take away some dreams of getting an easy and handy solution to this problem.

 

[Rich Z]

Yeah, there will be a lot of results that can lead to the target of the query, but I think perhaps a statistical percentage of probability would also be useful.

Honestly, I think many people really don't grasp what is involved in the production of multi homozygous and multi het animals. And LOTS of people come to me every year asking for cutting edge stuff and what I think would be cutting edge 5 years from now. So if someone has in mind of developing something unique based on combinations of known genes, there are obviously going to be some more efficient routes than others. Yeah, to some the path will be obvious, but to others, maybe not so apparent. I think this is absolutely the case that things are not at all intuitive nor clear when dealing with Ultra/Ultramel and derivatives. The path needed to get to some more complex combos using those sources may show that multiple parallel paths are needed in order to reach the desired goal.

 

[Ssthisto]

In other words, say someone claims that they have a Hypo Lavender Motley het Sunkissed Blood Red. What are the possibilities of the PARENTS needed to produce something like this?

Well, if it's absolutely het X and Y, then one parent had to be visual for each of those traits - either one was a visual X and the other was a visual Y, or one was a visual XY and the other was 'anything goes'. And if it's visual A, B and C, then the parents had to be at least het for those traits.

If I were going to guess what the possible parents of those offspring were... I'd have thought that you'd have a Hypo lavender bloodred het motley X Sunkissed Motley het hypo lavender... which is still pretty unlikely - it'd be a lot of work getting 'em to that point.

 

[paulh]

But suppose someone were interested in finally producing a Hypo Lavender Blood Red Caramel instead? What's the BEST road to take to reach that goal? :grin01:

hypo x hypo --> all hypos
hypo x het hypo --> 1/2 hypos
het hypo x het hypo --> 1/4 hypos

lavender x lavender --> all lavenders
lavender x het lavender --> 1/2 lavender
het lavender x het lavender --> 1/4 lavender

blood red x blood red --> all blood red
blood red x het blood red --> 1/2 blood red
het blood red x het blood red --> 1/4 blood red

caramel x caramel --> all caramel
caramel x het caramel --> 1/2 caramel
het caramel x het caramel --> 1/4 caramel

If the parents fit the genotypes in these matings, they can produce a hypo lavender blood red caramel. Of course, if both parents are heterozygous at all four loci, the odds of getting one is 1/4 * 1/4 * 1/4 * 1/4 = 1/128. In other words, the more loci are homozygous for the desired mutants, the better the odds of getting what you want.

 

[Austin M]

I agree it would be great to have a reverse genetic calculator. Some thing else that would be cool is getting persentages or odds from poss. hets.
Example: Which would be the quickest path to get lav bloods stripe?

Would the odds be better to breed a lav blood X het blood stripe and deal with F1 bloods 50% poss het stripe

or breed a lav blood X stripe and deal with triple hets? This is just an example.

It would be very cool to have a program that would statistically show odds for poss hets to determine quickest breeding routes.

 

[Menhir]

In other words, the more loci are homozygous for the desired mutants, the better the odds of getting what you want.

As said, breed two HypoLavenderBloodredStripes together and get the 100% odds.

Example: Which would be the quickest path to get lav bloods stripe?

Question, did you read my posting?

This is just an example.

Yes, but thats not a question of a reverse calculator, but a question of having two animals and calculating the odds. This is a different problem with different solutions possible.

It would be very cool to have a program that would statistically show odds for poss hets to determine quickest breeding routes.

Use my common calculator and divide the odds by 1/2 if one poss.het. is included and by 1/4 of there are two... and so on, and so on.

 

[ecreipeoj]

It seem as if I have seen photos you have posted of possible Lavender Caramels. Have you seen different results over the years in clutches that could have produced a Lavender Caramel? The reason I ask, is because of a question that was ask of Don S in a chat. Here is what he had to say about the combo many years ago. It seems as if yourself and others, may have had different results since this one.

Q: Have you seen any evidence in your own breedings or others that Caramel X Lav will be a worthwhile endeavor? Or is that top secret?

A: I produced my first F2 broods of lavender X butter seven years ago. I showed all the babies to Rich and he didn't see any obvious complex homos. I didn't either. There was surely masking of traits Vs stacking of traits and I never pursued it. Too much on my plate then AND now.

Q: Were they all amels?

A: I got all of it. Amels, caramels, butters, lavenders, and even hypos of all that mess. It was overwhelming. Opals too, of course.

Why do you think some Lavenders have normal eyes like your Avatar, and some have ruby eyes? Have you ever considered the possibility that “Mocha” and “Lavender” where alleles instead of the same gene? This is a very new concept, since we have discovered that Ultra is an allele with amel. This may answer some of the odd things we see in Lavenders such as our floating “hypo” like gene and the different results we see when combining Lavender and Caramel.

What if the breedings were actually, “Mocha” X Caramel, or “Lavender” X Caramel and the results were different? If they are alleles, we could see three different phenotypes in clutches when they are combined, which could be the answer to some of the odd results we see in the Lavender line, such as Don‘s “Hypos“, which I have seen in Lavender Lines that do not carry the Hypo gene. Is there just a “ruby eyed” gene or is it an indication of possible alleles?

I believe a breeding of a Hypo Lav Blood X Amber Blood would be the ticket. Hypo Bloods het Lav Car, would reduce the odds to 1 in 16 to produce a Hypo Lavender Caramel Blood.

 

[paulh]

Of course, if both parents are heterozygous at all four loci, the odds of getting one is 1/4 * 1/4 * 1/4 * 1/4 = 1/128.

Small mistake there. 1/128 is wrong; it should be 1/256.

I agree it would be great to have a reverse genetic calculator. Some thing else that would be cool is getting persentages or odds from poss. hets.
Example: Which would be the quickest path to get lav bloods stripe?

Would the odds be better to breed a lav blood X het blood stripe and deal with F1 bloods 50% poss het stripe

or breed a lav blood X stripe and deal with triple hets? This is just an example.

It would be very cool to have a program that would statistically show odds for poss hets to determine quickest breeding routes.

Go to Menhir's or one of the other calculators, and get the fraction of lav blood stripes from two matings: blood het lav het stripe x blood het lav het stripe and het blood het lav het stripe x het blood het lav het stripe. I calculate 1/16 for the first mating and 1/64 for the second. Of course, if either one of the snakes in the first mating is not het stripe, then there will be no lav blood stripes among the babies.

The secret to dealing with possible hets is to use possible hets at as few loci as possible and then raise a bunch of snakes so the ones that turn out to be duds do not derail the project. In other words, if you can only raise one male and one female, do not try to use possible hets. There is too good a chance of at least one being a dud. If you can raise 6 males and 6 females, then the risks of using possible hets for one locus are acceptable.