what determines if a gene is passed on?

[insane_alien]

Right i'm not so big in genetics but me and a few friends are making a robot which will be controlled by "genes" which are lines of code for it to display responses to stimuli etc. etc. there will be roughly 500(it can vary) of these per robot out of a selection of 2000 (thats all we can think of at the moment)

and since we know nothing of biology we thought we would come here to ask,

What determines if a gene is passed on during sexual reproduction?

Is there a reason for it or is it just random?

[ecoli]

well, considering 99.9% of humans genes are identical, I'd say it's mostly reason, because all humans are basically the same. Some genes are manifested physically (ie. hair color, eye colo, height). We acuire genes from both parents. What genes the offspring acuires from whom, is indeed random... but because the offspring is only gaining genes from two indivduals, the resulting gene combination is relatively predictable. (ie. two parents with brown hair will most likely have at least one child with brown hair.)

 

I haven't gone into more detail about dominant/recessive traits...how much did you want to know?

[insane_alien]

as much detail as you like its all good i just want to know how this little bit of genetics works.

[Mokele]

Basically, which gene is passed on depends on which sperm fertilizes which egg. This is random.*

 

Mokele

*It's not quite random for X versus Y, because the Y is lighter and thus sperm carrying it can move faster. But in all other cases, yes, it's totally random.

[ecoli]

*It's not quite random for X versus Y' date=' because the Y is lighter and thus sperm carrying it can move faster. But in all other cases, yes, it's totally random.[/quote']

 

then why are 52% of the population females.

[Yggdrasil]

Life expectancies may have something to do with it. Females tend to live longer than males (I don't realy know the biological basis for this, although the basis may involve social aspects). For example, in the US, women live about 5 years longer than men.

[Bluenoise]

Aparently lifestyle has alot to do with it as well. As men and womens lifestyles have been changing life expectancy for men in catching up to women. Both are increasing but the rate that it's increasing for men is almost double that for women. This phenomenon has been observed in most first world countries. I don't remember the link to the US study but here's a report from sweden, it says about the same thing in that respect.

http://www.socialstyrelsen.se/Publicerat/2001/2617/2001-111-2+Summary.htm

Not to say that there isn't some biological basis for it as well.

[Deathby]

Ok here's how it works. For every gene, there are two halves. eg. AA. When a normal cell turns into a sperm or egg cell, it doubles so it has AAAA, then splits into four cells, and each of those has an A.

 

But if there can be different forms of genes called alleles Eg. A and a. If a father has genotype Aa, then when he produced those 4 sperm cells, there will be 2 sperms with A and 2 with a, but its totally random which one is passed on to the child.

 

If a is a "crap" gene (not techical term) then if a child recieves it will die because of natural selection and that is how bad genes are weeded out. For the most part genes will not be weeded out during the reproduction process itself (that is once 2 mates have been selected). So I don't really see how you could use that to create a robot's coding. I assume you wanted to find out how you could get rid of erroneous coding using reproduction as a model.

[insane_alien]

Actually the erroneous coding(from random values generated by the program in "reproduction") tends to work itself out since the robot offspring that recieve it don't tend to live long enough to be able to reproduce or in some rare cases are sterile (the instruction telling it how to reproduce got screwed up). Much like in real evolution which is what i am looking for. Its not exact and sometimes the crap gene gets passed on and can screw the whole thing up where everything dies. Thanks for the help.

[Bluenoise]

Actually the erroneous coding(from random values generated by the program in "reproduction") tends to work itself out since the robot offspring that recieve it don't tend to live long enough to be able to reproduce or in some rare cases are sterile (the instruction telling it how to reproduce got screwed up). Much like in real evolution which is what i am looking for. Its not exact and sometimes the crap gene gets passed on and can screw the whole thing up where everything dies. Thanks for the help.

 

 

This is a very good representation of "genes" actually. Most of the replies that you have recieved have focused exclusivley on human genetics requiring two parents and sex. Mimicing humans makes little sense but to boost our own egos really. Different systems of genetics occur in large multicelular organism due to the difficulty in promoting diversity and evolution due to much longer generation times. For a robot that can demonstrate extreemly fast reproduction and generation times having two parental sources really isn't a necessity.

What you describe is very simular to genetic change in microbes, and I think it would be good to stick to it.

Heck you could even put some code into them that would occasionally allow them to transfer "genes" to another robot and see what effect that has and if this code itself would be promotted.

Good Job.

[insane_alien]

yeah the transfering of genes is how we simulate a virus we have the type of reproduction where the robt just spawns another like when a cell splits(whatever that process is called) with a few mutations sometimes. but we want the option for two (or more) robots get together transfer some information(mating i suppose you could call it) to make an offspring that is a mixture of the two plus a few random mutations as in the spawning method. maybe some kind of egg thats fertilized outside the body like fish. I don't know yet.

[Bluenoise]

Hmmm well I'd imagine the problem to be that if you want to have two copies of the each gene as in humans you'd have trouble trying to express both sets of code simultanously. How would you go about deciding if a trait is dominat or recessive or co-dominant etc...?

[insane_alien]

well i had an idea to give each gene a variable for dominance like

dominant=1 ; //gene is dominant

dominant=0 ; //gene is recessive

but the problem is if i get a dominant gene that does one thing and a dominant gene that is incompatible with the first gene(these are rare but cause program crashes)

multiple instances of a gene aren't much of a problem but can lead to the robot "dying" if it is a rotation function as it can't aim as well to get food but generally its ok. whats co-dominant?

[insane_alien]

i think i might implement a gender with "chromosones" just the two types though Y will be male and X female (its easier to sort out the behaviours than with XX and XY).

by the way i'm taking it as dominant genes are passed on and recessive genes are not, is this right?

[Skye]

by the way i'm taking it as dominant genes are passed on and recessive genes are not, is this right?

No. It has no impact on whether a gene is passed on or not, only on whether a gene is expressed in the phenotype.

 

If gene A is dominant with respect to gene B, then the A phenotype will result and the B phenotype won't. B is then recessive to A. But another gene C might be dominant to gene A, and A is the recessive to C. So A can be dominant to one gene and recessive to another. Dominance and recessiveness aren't properties of the genes themselves, but of the relationships between the genes.

[Mokele]

by the way i'm taking it as dominant genes are passed on and recessive genes are not, is this right?

 

Not quite. I'll explain.

 

Most animals are diploid. This means they inherit one set of chromosomes from their mother, and one from their father. Each set of these chromosomes contains all the genes an organism needs, thus a fertilized egg have two copies of every gene, one from mom and one from dad. For instance, you have one gene for alpha-hemoglobin from mom, and one gene for the same protien from dad.

 

Genes, like computer programs, come in versions, called alleles. While both your mom and dad's genes make alpha-hemoglobin, they might be different alleles, different "versions". Each allele (usually) makes the same overall protien, but there are subtle differences in the final form. The alpha-hemoglobin from your mom's gene might look and behave slightly differently from the alpha-hemoglobin from your dad's gene.

 

These two alleles can interact. Sometimes, one allele is "dominant". This means, no matter what the other allele is, only the "Dominant" version is expressed. Others are "recessive", and can *only* appear if both copies (from mom and dad) are the same recessive allele. In Co-dominant, individuals who have one copy of each type express an intermediate phenotype.

 

As for passing on to offspring, it's a craps shoot. When eggs or sperm are formed, they each only get one set of chromosomes, half of the organism's set. A sperm will have either mom's or dad's allele for a particular gene, but not both. Another gene might have the other parent's allele. There's some non-randomness, but that's rather more complex than I think you want. Which the offspring gets basically depends on which sperm gets to which egg, which is random (more or less).

 

To put it in programming terms, think of an array. Each position is a "gene" (or behavior). The elements in each array can be a series of values (like alleles of a gene) for different forms of that behavior. An adult organism has 2 arrays, and when a particular array element is called on, the values in that element of both arrays interact via dominance, recessiveness, co-dominance, etc to produce an effect. When that organism reproduces, it produces gametes with just one array, and the elements of that array are filled randomly from the elements of the two arrays the adult has. The gametes combine, and now the new program has 2 arrays again. Repeat.

 

Does that help?

 

Mokele

[insane_alien]

i'm going to have to take a while to think this through properly i think i get the general idea though. i get back to you on this.