Naked Science Forum
Life Sciences => Cells, Microbes & Viruses => Topic started by: mpt-matthew on 12/01/2011 14:12:05
-
Excluding crossing over during synapsis there is 8 million (2²³) combinations of chromosomes from a mother and farther.
There are 6,892,900,000 humans on earth. There is an average 2.49 babies per family.
Presuming around 90% if couples have babies, would this not meen that there is a strong possibility that one family on earth will have 2 babies on separate occasions that have almost identical genetics (i dont meen just look similar)?
Anyone good at maths can calculate (predict) the number of families on earth that will have identical twins on separate occasions.
thanks
------------------------------------------------
Someone answered the following:
Good question!
I never like to use absolute statements but this particular example really couldn't happen. The chances of the same man producing sperm with the identical genetic information is not just unlikely, but impossible. Sperm are not exact copies of each other and no two are identical. A man will never, in his lifetime, produce two sperm that are genetically identical. Combined with the fact that no two eggs in a female are identical and, really, the only way identical twins can be born is from a single fertilized egg that splits.
That said, identical twins could potentially be born many years apart but only with the help of modern fertility procedures. Theoretically, one embryo of a split embryo pair could be implanted into a woman now and the other frozen. Then years later, the second embryo could be implanted and, viola, identical twins with different birthdays. This has happened with fraternal "twins" recently. It should be feasible with identical twins.
--------------------------------------------------
I replied:
So you say both sperm aren't ever genetically identical. Im taking this is due to crossing over.
I still cant see how you can say never? It is highly improbable, 8 million possibilities + crossing over.
But considering the number of men (billions) and the number of sperm each man produces, it is surely mathematically highly probable this could happen.
Just like winning the lottery, it is unlikely you will win (i think around 1/80million to win the euromillion jackpot) but still people do win (as there are a lot of people).
---------------------------------------------------
Thanks, any maths etc.
Mod edit - formatted subject as a question - please do this to help keep the forum tidy and easier to navigate. Thanks.
-
In humans there are generally 1-3 cross overs per chromosome!
So that makes it even more unlikely. - nearly impossible, but still possible.
-
Matthew - your figure of 2^23...that would be correct if chromosomes were either or choices. But they are not each chromosome varies internally to a greater or lesser extent. The number of protein encoding genes is probably in the region of 15-30,000 - with some reckoning even higher. And even then these are not always on/off choices that could be modelled as 2^x.
The number of possible human combinations breaks my calculator... One of the geneticists on the site might have a better idea
-
Matthew - your figure of 2^23...that would be correct if chromosomes were either or choices. But they are not each chromosome varies internally to a greater or lesser extent.
Why aren't chromosomes an either or choice, is that just crossing over?
I understand now, due to the high amount of crossing over that the chances of this happening are very very slim.
BUT (regarding the quote)
Surely during meiosis the homologous chromosomes (2 chromosomes) line up, they are then split (after crossing over). so surely that is a 2²³ combinations (excluding crossing over).
each chromosome can either be in half 1 or 2, and there are 23, which ≈ 8 million.
-
thats the rough number of combinations any two parents can have - ie you get half as a selection from your fathers and half as a selection from your mothers. But my mother and your mother have different selections to choose from - although they did both call their sons Matthew!
-
Matthew - your figure of 2^23...that would be correct if chromosomes were either or choices. But they are not each chromosome varies internally to a greater or lesser extent.
Why aren't chromosomes an either or choice, is that just crossing over?
I understand now, due to the high amount of crossing over that the chances of this happening are very very slim.
BUT (regarding the quote)
Surely during meiosis the homologous chromosomes (2 chromosomes) line up, they are then split (after crossing over). so surely that is a 2²³ combinations (excluding crossing over).
each chromosome can either be in half 1 or 2, and there are 23, which ≈ 8 million.
Have you factored in alleles into it?
-
Have you factored in alleles into it?
Alleles are just an essences of the genetic code. What it codes for. Its all contained on the chromosomes, so yes.
-
Have you factored in alleles into it?
Alleles are just an essences of the genetic code. What it codes for. Its all contained on the chromosomes, so yes.
Errr not entirely no. If you are calculating the odds of identical genomes occurring then alleles feature heavily into that. They are not just simply an essence.
And then you have to consider SNPs...... I am no maths guru, but I would guess the calculations are impossible to do.
-
thats the rough number of combinations any two parents can have - ie you get half as a selection from your fathers and half as a selection from your mothers. But my mother and your mother have different selections to choose from - although they did both call their sons Matthew!
I think you thought i was asking something else, i was talking about per couple, not for me to be identical to another baby from another couple.
Because the average family size is above 1, each couple on earth has aprox 1 go at creating an identical twin using the same combination of chromosomes as the previous child.
Though it is unlikely...
Corrected calculations (making A LOT of assumptions - no crossing over) mathematical model:
In fact i am correcting my self, (as i will show below*) it is 2²³/2=4million (excluding crossing over).
There is a 1 in 4million chance 2 gametes from the same person will be the same (excluding crossing over).
Fertilisation takes place, probability of the same gametes meeting twice = Chance of getting same male gamete 1/4m, and same female, 1/4m
Total = 1 in 17 trillion, 592 billion, 186 million, 44 thousand and 416 PER COUPLE
Standardised as a rate of 19 births per year per 1000 people that makes 130965100 births pear year for the whole population.
Because the average family size is 2< half of the babies will be 2nd children, and hence be a possible twin (some will be 4th child's, making it a higher probability, some will be 1st excluding them)
If the current birthrate and population remains the same, it will take 134,327 years before 2 combinations the same will appear (by probability).
Obviously it could happen at any time (you cant predict where in that time scale it could happen).
It could have already happened, it may never happen.
Obviously none of this takes into consideration crossing over - this makes the maths much more complicated.
Obviously decreases the odds millions fold.
Also the average family size could be taken more into consideration (if average family sizes are high, then it will take less time, if they are low it wil slow it down).
The only time it would ever mathematically reach a 0 probability is if every family only had 1 baby (then you could never have twins).
In conclusions, highly highly unlikely, but still with some massive numbers, possible.
*
To look at it simply, if you have 2 homologous pairs of chromosomes (4 in total), and they line up in metaphase I. There are 2^4/2 combinations they can take.
(its hard to explain without a diagram), they could be on the same side or different sides (2 combinations). look: http://staff.jccc.net/pdecell/celldivision/images/metaphase1.gif
These are on different sides, the other option is they are on the same side.
If you increase the number to 3 pairs, there are 4 combinations. Remember the reverse combinations (what the /2 takes care of) dont count, as they are the same as the other combination).
In humans, there are 23 pairs, so thats 4million combinations.
-
Errr not entirely no. If you are calculating the odds of identical genomes occurring then alleles feature heavily into that. They are not just simply an essence.
And then you have to consider SNPs...... I am no maths guru, but I would guess the calculations are impossible to do.
I think you have misunderstood what i am trying to calculate.
In my calculation, variation between each couple is not important. (im not trying to calculate the same genome occurring through different couples but the same one).
i.e. 2 sisters having the same genome but born on different occasions.
You are correct, it is impossible to calculate, but the same goes for weather, or the decaying of an atom. You can still make a mathematical model which could predict an outcome based on variants.
As i stated, one variant i have not included is crossing over (though if you were good at maths you could factor this in). The more things you factor in, the better the model, the better the prediction. But you can never make a perfect model.
-
Ok...
I think your numbers are off. Here is how I'm calculating it.
If you have 23 pairs of chromosomes... from the father... of which you have to select half of the chromosomes... you can think of it as binary numbers. You get either chromosome 0 or 1. So, there are 223 combinations.
Same from the mother... so she gives 223 combinations.
The M/F choice is just one of the choices by the father, and not important for the calculations.
So...
you get 223 * 223 = 2(23+23) = 246
I.E. It would be the same as calculating it with 46 chromosome pairs.
246 = 70,368,744,177,664 (70 trillion choices).
Now, consider the world population.
6,893,132,254 (just short of 7 billion).
If you could simplify it down to sibling pairs with the same mother and father... and just 2 children per couple... is that an "Ideal World"?
That would cut your numbers in half... and you would have about 3,446,566,127 pairs.
I think you can just divide the two numbers....
70 quadrillion / 3 billion ==> about 1:20,417 chance that there would be a single sibling pair alive with identical chromosomes alive today.
I'm probably off by a factor of 2 or 4 somewhere, but the chance is still substantially less than 1 that there would be an identical twin pair alive today.
The "Crossover" issue becomes much more complicated.
While there may be some common crossovers, the spice could occur essentially anywhere on the chromosome, so you get essentially an infinite number of actual combinations.
---------------------------------------------
(oops, I said quadrillion... but read the number wrong... I meant trillion. No other changes made).
-
Ok...
I think your numbers are off. Here is how I'm calculating it.
If you have 23 pairs of chromosomes... from the father... of which you have to select half of the chromosomes... you can think of it as binary numbers. You get either chromosome 0 or 1. So, there are 223 combinations.
My last calculation resembles something similar to this.
2 differences.
1, you calculated the sibling chance by halving the whole population, which makes perfect sense! thanks
2, you use 2²³, not 2²³/2 which i used in my last calculation.
I think 2²³/2 is correct now (after some thinking), unlike my originally stated 2²³.
This is because in a binary system, each half (i.e. left and right side, or 0 and 1) is different, i.e. if you have a 2² binary system, there are 4 combinations, both right, one right, other one right, none right (if that makes sense), or 01, 10, 00, 11. But in a cell, each option is the same, so both right (11) and both left (00) in the end produce the same collection of gametes.
Look at this pic: http://staff.jccc.net/pdecell/celldivision/images/metaphase1.gif this is one right, if it was other one right you would still get exactly the same 4 gametes at the end of meiosis.
Therefore there is a total of 2 possible gamete options which is 2²/2. The same goes for a 2³ system, there are 4 options.
PS Im not sure how you would do it, but your calculation doesn't give a 0 probability if the average number of children per family was 1 (which it should be). Im not sure where you would incorporate this.
-
PS Im not sure how you would do it, but your calculation doesn't give a 0 probability if the average number of children per family was 1 (which it should be). Im not sure where you would incorporate this.
Cousins marrying Cousins?
Oedipus?
A stable (or healthy slowly declining) population is 1 kid per parent, or 2 kids per couple (excluding Oedipus, of course which would change your averages).
Anyway, the diffusion of genes in the population is so rapid that the probability of "twins" in the general population drops to virtually zero. The problem is that in the general population, you don't have 23 or 46 chromosomes... but virtually every chromosome is unique due to random variations and cross-overs. The only way to do the calculation would be in a very tight tribal society... and still you'd need to enforce no mutations or crossovers. Did we talk about spontaneous mutations? I think the spontaneous mutation rate is on the order of a few hundred base pairs per generation.
I'm still thinking the chances are 1:246, rather than 1:223. (perhaps dividing by a factor of 2).
As I believe it would be the same to line up 23 maternal and 23 paternal chromosomes and dividing them... vs lining up all 46 chromosomes (keeping maternal and paternal pairs separate)... and dividing them.
Or, as I discussed earlier, giving each maternal chromosome a 0/1 bit, and each paternal chromosome a 0/1 bit, thus 46 bits total.
Now...
An interesting calculation would be the probability of getting an "anti-twin" (without scientifically creating it).
During telophase, every egg and every sperm has a pair with all the opposite chromosomes. The male sperm, without fertilization treatment, of course, would die within 9 months. But, perhaps it could happen with fraternal twins.
Keep in mind that through fertilization treatment, it is relatively easy to create identical twins, either both born at the same time, or later dates.
-
ou are correct, it is impossible to calculate, but the same goes for weather, or the decaying of an atom. You can still make a mathematical model which could predict an outcome based on variants.
As i stated, one variant i have not included is crossing over (though if you were good at maths you could factor this in). The more things you factor in, the better the model, the better the prediction. But you can never make a perfect model.
With you now.
Even so, you still have 4 possible alleles of each gene to consider, and SNPs.
You could make a mathematical model for most things, but that does not mean you can exclude factors that change the numbers that you crunch.
-
PS Im not sure how you would do it, but your calculation doesn't give a 0 probability if the average number of children per family was 1 (which it should be). Im not sure where you would incorporate this.
Cousins marrying Cousins?
Oedipus?
A stable (or healthy slowly declining) population is 1 kid per parent, or 2 kids per couple (excluding Oedipus, of course which would change your averages).
Anyway, the diffusion of genes in the population is so rapid that the probability of "twins" in the general population drops to virtually zero. The problem is that in the general population, you don't have 23 or 46 chromosomes... but virtually every chromosome is unique due to random variations and cross-overs. The only way to do the calculation would be in a very tight tribal society... and still you'd need to enforce no mutations or crossovers. Did we talk about spontaneous mutations? I think the spontaneous mutation rate is on the order of a few hundred base pairs per generation.
I'm still thinking the chances are 1:246, rather than 1:223. (perhaps dividing by a factor of 2).
As I believe it would be the same to line up 23 maternal and 23 paternal chromosomes and dividing them... vs lining up all 46 chromosomes (keeping maternal and paternal pairs separate)... and dividing them.
Or, as I discussed earlier, giving each maternal chromosome a 0/1 bit, and each paternal chromosome a 0/1 bit, thus 46 bits total.
Now...
An interesting calculation would be the probability of getting an "anti-twin" (without scientifically creating it).
During telophase, every egg and every sperm has a pair with all the opposite chromosomes. The male sperm, without fertilization treatment, of course, would die within 9 months. But, perhaps it could happen with fraternal twins.
Keep in mind that through fertilization treatment, it is relatively easy to create identical twins, either both born at the same time, or later dates.
Hmm, interesting.
I was trying to calculate the chances of the same couple conceiving on different occasions genetically (or almost) identical twins through the coming together of the same combinations of chromosomes.
The 2 main areas where genes are mixed up is random chromosome assortment, and random sperm/egg pairing.
Each can be calculated.
Regarding the probability of 1 gamete been the same as another: 0.5(246) or 0.5(223)
Im taking it you understand that you have to divide it by 2, as a 11 00 combination and a 00 11 combination produce the same gametes int he end (though register as 2 separate configurations in binary).
Now why it is 23...
This splitting takes place during Metaphase I, which is where there are 23 lines of chromosomes, in each line there is 2 chromosome (homologous).
It looks like this (http://www.biologycorner.com/resources/metaphaseI-2.gif (http://www.biologycorner.com/resources/metaphaseI-2.gif)) but immagine that there are 23 lines instead of 4. This gives a total of 46 chromosomes.
Each line splits like so http://www.biologycorner.com/resources/anaphaseI-2.gif (http://www.biologycorner.com/resources/anaphaseI-2.gif)
Each line the mother (1) or farther (0) homologous chromosome can be on the right (R) or left (L)
so:
lets say you only have 3 lines...
L R
0 1
0 1
0 1
L R
0 1
1 0
0 1
L R
1 0
1 0
0 1
L R
0 1
1 0
1 0
This is 4 combinations 0.5(23) from a total of 6 chromosomes (3 homologous pairs), there are no more combinations which will yield a different set of gametes.
This makes the chance of any perticular gene combination, ( ½ (2 23 ) ) 2= 1 in 1.76X10 13
I hope this concludes how the calculation of the chance of the same gamete is.
If the average babies per couple was 1, i dont think you could have cousins, therefore 0 probability. But this is not the case, as the population (as you said) will eventually decline and die out.
By anti-twin i presume you meen another sibling from the same couple with the opposite chromosome combination of you.
This would be the same probability as creating an exact twin - the probability of getting ANY gene combination is 1 in 1.76X10 13.
This probability can then be applied to the population to model what the chances of a identical or antitwin birth are.
-
Ahhh.... The alleles.
Except for the X-Chromosome, the alleles are not chromosome dependent, but rather gene dependent.
So, you could have some dominant genes on a Paternal Chromosome other dominant genes on the corresponding Maternal Chromosome. I suppose then you could only look at the transmission of dominant genes for "phenotype twins", especially if you are looking for non-related phenotype twins. But, then it would all depend on the resolution of what you call a "twin".
In men, of course, the X Chromosome is unpaired.
In women, it is a bit more unique because there is something called X Chromosome Inactivation.
http://en.wikipedia.org/wiki/X-inactivation
So, in a sense you could ignore the inactive X chromosome in your phenotype comparison... except that the inactivation isn't complete.
-
With you now.
Even so, you still have 4 possible alleles of each gene to consider, and SNPs.
You could make a mathematical model for most things, but that does not mean you can exclude factors that change the numbers that you crunch.
No,
Each human has around 4 million different variations of gamete.
Each allele and SNP is a varient in the genetic information. As each cell in the human body is the same, all the sperm a man produces during his life, and all the eggs that a woman is born with all have a chromosome makeup of 1 in 4million.
Do you understand how meiosis works?
The DNA (which contains the alleles and SNPs) is contained on the chromosome.
If you fertilise an egg with a sperm, then repeat it with a sperm and egg with identical chromosome configuration you will get an identical zygote (except for crossing over, as i mentioned).
Alleles are just a feature of the genetic code, in an identical twin even the non-coding areas of DNA is the same.
As stated by Clifford K, you could create a phenotypical twin, but then that would be subjective to the amount of phenotypical similarity.
Hope this helps. If you still disagree, please explain yourself more.
-
I'm not sure about where you get the half.
Consider the 3 pairs of chromosomes... A, B, & C.
Father (paternal) has the genes, 3 pairs (6 chromosomes total) (A0pA1pB0pB1pC0pC1p)
A0p, A1p
B0p, B1p
C0p, C1p
Mother (maternal) has the genes, 3 pairs (6 chromosomes total) (A0mA1mB0mB1mC0mC1m)
A0m, A1m
B0m, B1m
C0m, C1m
Father can create the possible haploid combinations (23 = 8 total)
A0pB0pC0p
A0pB0pC1p
A0pB1pC0p
A0pB1pC1p
A1pB0pC0p
A1pB0pC1p
A1pB1pC0p
A1pB1pC1p
Mother can produce the haploid combinations: (23 = 8 total)
A0mB0mC0m
A0mB0mC1m
A0mB1mC0m
A0mB1mC1m
A1mB0mC0m
A1mB0mC1m
A1mB1mC0m
A1mB1mC1m
So, now you get 8 paternal haploid combinations, and 8 maternal haploid combinations... 23 * 23 = 26 = 8*8 = 64 diploid combinations.
Do I need to list out the entire list? (written in 2 separate columns to save space). 64 total.
A0mB0mC0mA0pB0pC0p A1mB0mC0mA0pB0pC0p
A0mB0mC0mA0pB0pC1p A1mB0mC0mA0pB0pC1p
A0mB0mC0mA0pB1pC0p A1mB0mC0mA0pB1pC0p
A0mB0mC0mA0pB1pC1p A1mB0mC0mA0pB1pC1p
A0mB0mC0mA1pB0pC0p A1mB0mC0mA1pB0pC0p
A0mB0mC0mA1pB0pC1p A1mB0mC0mA1pB0pC1p
A0mB0mC0mA1pB1pC0p A1mB0mC0mA1pB1pC0p
A0mB0mC0mA1pB1pC1p A1mB0mC0mA1pB1pC1p
A0mB0mC1mA0pB0pC0p A1mB0mC1mA0pB0pC0p
A0mB0mC1mA0pB0pC1p A1mB0mC1mA0pB0pC1p
A0mB0mC1mA0pB1pC0p A1mB0mC1mA0pB1pC0p
A0mB0mC1mA0pB1pC1p A1mB0mC1mA0pB1pC1p
A0mB0mC1mA1pB0pC0p A1mB0mC1mA1pB0pC0p
A0mB0mC1mA1pB0pC1p A1mB0mC1mA1pB0pC1p
A0mB0mC1mA1pB1pC0p A1mB0mC1mA1pB1pC0p
A0mB0mC1mA1pB1pC1p A1mB0mC1mA1pB1pC1p
A0mB1mC0mA0pB0pC0p A1mB1mC0mA0pB0pC0p
A0mB1mC0mA0pB0pC1p A1mB1mC0mA0pB0pC1p
A0mB1mC0mA0pB1pC0p A1mB1mC0mA0pB1pC0p
A0mB1mC0mA0pB1pC1p A1mB1mC0mA0pB1pC1p
A0mB1mC0mA1pB0pC0p A1mB1mC0mA1pB0pC0p
A0mB1mC0mA1pB0pC1p A1mB1mC0mA1pB0pC1p
A0mB1mC0mA1pB1pC0p A1mB1mC0mA1pB1pC0p
A0mB1mC0mA1pB1pC1p A1mB1mC0mA1pB1pC1p
A0mB1mC1mA0pB0pC0p A1mB1mC1mA0pB0pC0p
A0mB1mC1mA0pB0pC1p A1mB1mC1mA0pB0pC1p
A0mB1mC1mA0pB1pC0p A1mB1mC1mA0pB1pC0p
A0mB1mC1mA0pB1pC1p A1mB1mC1mA0pB1pC1p
A0mB1mC1mA1pB0pC0p A1mB1mC1mA1pB0pC0p
A0mB1mC1mA1pB0pC1p A1mB1mC1mA1pB0pC1p
A0mB1mC1mA1pB1pC0p A1mB1mC1mA1pB1pC0p
A0mB1mC1mA1pB1pC1p A1mB1mC1mA1pB1pC1p
So, extrapolating to 23 pairs of chromosomes, you get 223*223 = 246 = 1 in 70,368,744,177,664 chance.
If the average babies per couple was 1, i dont think you could have cousins, therefore 0 probability. But this is not the case, as the population (as you said) will eventually decline and die out.
Good point about no cousins, at least back to the point where one started the 1 child policy. So you're just left with Oedipus.
I have always wondered why China's population is still increasing with a 1 child policy.
By anti-twin i presume you mean another sibling from the same couple with the opposite chromosome combination of you.
This would be the same probability as creating an exact twin - the probability of getting ANY gene combination is 1 in 1.76X10 13.
True... except there will be 1 sperm, and 1 egg that has the exact opposite genes.
The likelihood of the second egg getting fertilized either at the same time (fraternal twins), or during successive pregnancies is non-zero. So, if fertilization is "random", with fraternal twins, it would depend on the number of sperm per ejaculate, and would be around 1 in 280 Million (assuming both sperm are viable in the same ejaculate).
For successive pregnancies, it depends on the chance of that second egg getting fertilized times the probability of the "anti-twin" sperm, which as mentioned, with no crossovers, it is actually a smaller number 1 in 223 for the men (1 in 8,388,608), but in reality is a more rare case due to the crossovers.
Without getting that second egg... the obviously the number is the same as the one calculated above... 1 in 70 trillion.
(did I say quadrillion earlier... I guess I meant trillion).
I wonder if the female's egg would account for why many fraternal twins have a high degree of differences.
-------------------------------------
-
I'm not sure about where you get the half.
I have always wondered why China's population is still increasing with a 1 child policy.
True... except there will be 1 sperm, and 1 egg that has the exact opposite genes.
The likelihood of the second egg getting fertilized either at the same time (fraternal twins), or during successive pregnancies is non-zero. So, if fertilization is "random", with fraternal twins, it would depend on the number of sperm per ejaculate, and would be around 1 in 280 Million (assuming both sperm are viable in the same ejaculate).
For successive pregnancies, it depends on the chance of that second egg getting fertilized times the probability of the "anti-twin" sperm, which as mentioned, with no crossovers, it is actually a smaller number 1 in 223 for the men (1 in 8,388,608), but in reality is a more rare case due to the crossovers.
Without getting that second egg... the obviously the number is the same as the one calculated above... 1 in 70 trillion.
Yes, you are correct. The half shouldn't be there.
Thanks for the help with all the calculations - so the chances are around 1 in 70 trillion. Still not 0 (but obviously this doesn't factor in crossing over).
And yes, i was also wondering about chinas 1 child policy and the increase in population.
Thanks :)
-
Do you understand how meiosis works?
The DNA (which contains the alleles and SNPs) is contained on the chromosome.
Meiosis, hmmmm now I wonder what that is... [::)]
Do you understand the word patronising?
Hope this helps. If you still disagree, please explain yourself more.
Not particularly because you are concentrating on making the numbers crunch whereas I am basing it on what actually happens and what is feasible.
[::)] [::)] [::)]
-
Not particularly because you are concentrating on making the numbers crunch whereas I am basing it on what actually happens and what is feasible.
[::)] [::)] [::)]
Sorry if i appeared patronising
I am genuinely interested in why i am not understand what you are saying.
Please explain how alleles and SNPs would change from one conception to another, if both sperm and egg had the same chromosomal configuration?
Alleles, genes, SNPs etc are carried on set chromosomes, so if you 2 of the same gamete as long as crossing over doesn't take place (i know it does, but we factored this our for a rough calculation), then there would not be a difference.
-
Even so, you still have 4 possible alleles of each gene to consider, and SNPs.
Here is why i think alleles and SNPs dont need to be taken into consideration in this calculation.
Firstly for anyone who is reading this back and might not know, and allele is "either of a pair (or series) of alternative forms of a gene that can occupy the same locus on a particular chromosome and that control the same character; "some alleles are dominant over others" ".
I will use an example to help, lets say eye colour.
I have 2 alleles for eye colour, lets say Bb (heterozygous, blue/brown), brown from my mum and blue from my dad.
The genetic information for eye colour lets say is stored on chromosome homologous pair 15. So on one of the chromosomes i have the genetic info for brown, and on the other blue.
Brown is dominant so i will exhibit this phenotype.
When meiosis takes place, during metaphase I the homologous chromosomes line up down the metaphase plate and they are split into 2 haploid cells.
How the chromosomes mix up is random, BUT the opposite homologous pair goes to the other cell. So if my brown allele goes to the right cell, the blue allele will go to the left cell.
These cells then divide again to produce 4 gametes.
We where calculating the chance that 2 specific gametes will come together, i.e. the chance of having the same genes form an embryo.
Because there are 23 homologous chromosomes, this gives a 1 in (2²³)² probability that a specific sperm egg combinations will form.
If 2 babies where born from 2 sperm and eggs which had the same chromosomal configuration then they would be identical, except for differences brought about by crossing over (which could be different). The alleles and SNPs they had would be the same, because these are genetic features stored on the chromosomes, and they have the same chromosomes.
Hope this makes sense, if you spot something that is not clear, or you think is incorrect, please point it out.