Quantitative Genetics, Breeding Mathematics

[titoon29]

Hi Everybody,

Few weeks ago, with a friend grower, we've discussed on how mathematics could be applied in Cannabis Breeding (we re going crazy learning how maths can explain a lot of things lol )...
So since we've found that there was some breeding mathematics, we'll need to fully understand them before anything.... After that we'll have to find some datas about cannabis breeding, but that's another story....
I was wondering if anyone has tought about that, and if he could give me any advice on what are the interesting books i should pick up...
Books i ve found :
http://books.google.fr/books?q=bree...=bb0Ta3UWV1&sa=X&oi=print&ct=title&cad=legacy

Any idea if any breeder has done some research about it ?

That s going to be a crazy quest i think ^^
Hope you'll be able to understand my bad english

Thanks !

 

[Grasso]

Hello,

sexual propagation means that genes are mixed at random. Therefore you should have a look at probability theory.

Uli

 

[titoon29]

Hi Grasso, thank for your reply....

In mathematics, "random" doesn't mean anything my friend always says ^^
I understand what you mean, i don't hope to came with an exact formula to give the exact dad and mother for an awesome cross...
It s just that, when you use maths, you could greatly improve breeding skills, as it shows you every details that's important or if some detail's missing.
In term of phenotypic variation, it can help, with a some data from previous grows, to find the good mother and father... That's what breeders do, and i have a great respect for them, some really have great selection methods, just want to have a different approach of breedingn more cartesian.

 

[Dr.NO]

1 male plant plus 1 female plant equals a bunch of seeds.

 

[titoon29]

I agree doctor No

But if everyone did that you wouldn't get so much great seeds....
And when i'm talking about breeding, it's breeding with a lot of Parents....

What i would like to try to find out on how genetics are sent from parents to children, with statitistics...

There is way too much going on in the Cannabis genome for you to be able to pick the best male using your senses to evaluate his traits before you breed him. Its like an iceberg. There is so much going on below the surface that you just cant know whats there without running into it.

Well that why maths can, with few datas, simplify the genome, and try to explain how genes are sent when pollinating... But that is far from what i expect to understand.
At a lower stage, using some maths could really help find out where are the problems or what could be improved in actual breeding methods, but for that you need data...
i think i'll need a lot of time to start to understand and make conclusion, as there are breeders who have a life full of experience... But i think it's worth it.

You should look the intros of the books from the link i sent, they explain very well the problem... As it seems we have a few genetic mathematicals models, there is a plant model which help them increasing the crop from corn. i'm pretty sure there s some people who have done research on this way of beeding, as i know few people fond of math who are smoking and growing...

Again, hope my english s not too bad, i have some hard time understanding genetic's english...

Thank

 

[PazVerdeRadical]

titoon29,
if we consider that we have to run into most genetic traits in a plant before we know what they are there,
you cannot expect to be able to predict them with a formula; however, you can predict how genetic traits you have already studied and quantified and qualified may turn out if bred this or that way with yet another set of genetic traits that you already know about.
i don't really see where u r going...

peace.

 

[Fat Albert]

Probability theory could indeed be useful, when used to determine possible genetic variants due to mutation and other unexpected influences, but when it comes to incomplete dominance, heterozygosity, and other aspects of individual genetic expression, I'm afraid that our current science limits us to just growing 'em out and seeing what happens. Math cannot accurately predict a phenomenon when we don't know if we've got all the inputs right (or even all the inputs, for that matter).

That being said, I think it's a fascinating subject!

Fat A

 

[subfloral]

when it comes to biological systems, there is no such thing as random... if we call a process random, then most likely it is because our understanding of it is incomplete.... researchers are discovering before unknown subtle biochemical interactions every day

mathematics have been invaluable in formulating theoretical frameworks for understanding living systems, and can be extremely helpful in understanding genetics
but
until our understanding of these interactions is complete, then our math will be a little off

 

[Haps]

A + B = The math is around here somewhere, read some of DJ Short's stuff if you are looking for methodology. The real deal is when you feel the equation in the dance of your growing life. The harmony of possibilities and the dischords of poor phenotypes, against the glorious stone of your own magical creations. Tune in to your plants and they will lead you.
H

 

[- ezra -]

Well, its a sound idea titoon. And indeed mathematics can be used to predict genotype ratios for breeding. Simple calculations like punnet squares and such might be helpful when looking at simple things like single locus traits with few alleles.

But the problem is that the mathematics can get so complex that it really is not practical to work with IMO. When you have traits which are under multiple gene control, with many alleles involved, and various dominance relationships, and then you want to calculate ratios for multiple traits as well, it gets very complex I would say.

On top of the mathematical complexity, in order to be able to work with mathematical models or equations, you need data about the traits you are interested in to be able to perform the calculations. That would mean you would need to know how many loci are controlling each trait, how many alleles are there in play, what are the dominance relationships. You need to know if the plants your working with are homozygous or heterozygous for the various alleles, etc...

In practice, it actually makes more sense to use a trial and error type approach combined with simple theoretical guidelines IMO.

peace, Ez