PDA

View Full Version : Question about DNA samples



Project nachonaco
07-02-2010, 11:46 PM
1. What's the easiest way to get a DNA sample from a two-year-old without making her upset?

2. When comparing her DNA, (let's say the place she was found at demands DNA samples from every employee), would her aunt show up in the list of DNA matches by approximation?

PeterL
07-02-2010, 11:56 PM
The common way to take DNA samples is a swab of the inside of the mouth, and I think that would work for a two year old.

An aunt could be a close match or rather distant. The child's mother's sister would have the same X chromosome, but all other genes could be different. Beyond that it is impossible to say with any certainty, because we don't know whether the father might have been a first cousin, or what.

Project nachonaco
07-03-2010, 12:00 AM
The child's mother is the aunt's sister, if that matters.

PeterL
07-03-2010, 12:45 AM
The child's mother is the aunt's sister, if that matters.

That means that they have the same X chromosome, but it tells nothing beyond that, because all other the father and from the DNA that the two sisters do not share. The odds are that they would share about 25% of their genes, but that's just the midpoint of the probability curve. It is conceivable that they would have all of the same genes or as little as one shared gene.

You might want to look into genetics to see how the probabilities run.

Drachen Jager
07-03-2010, 01:43 AM
That means that they have the same X chromosome, but it tells nothing beyond that, because all other the father and from the DNA that the two sisters do not share. The odds are that they would share about 25% of their genes, but that's just the midpoint of the probability curve. It is conceivable that they would have all of the same genes or as little as one shared gene.

You might want to look into genetics to see how the probabilities run.

That's not true at all. Familial traits will show up in both aunt and niece, they can track family trees back through many generations.

With strictly female lineage it's even easier because only women pass on the mitochondrial DNA.

see http://genealogy.about.com/cs/geneticgenealogy/a/dna_tests.htm

Furthermore, ALL humans share 99.9% of their DNA. Something with only 25% of a human's DNA wouldn't even be an animal.

ladyleeona
07-03-2010, 03:17 AM
That's not true at all. Familial traits will show up in both aunt and niece, they can track family trees back through many generations.

With strictly female lineage it's even easier because only women pass on the mitochondrial DNA.

see http://genealogy.about.com/cs/geneticgenealogy/a/dna_tests.htm

Furthermore, ALL humans share 99.9% of their DNA. Something with only 25% of a human's DNA wouldn't even be an animal.

This is what I was about to say--isolation of mitochondrial DNA from a cheek swab is possible. Like Drachen said though, mitochondrial DNA will only prove that the aunt is related to the child's mother, and not any paternal DNA inheritance.

Furthermore, while both the females would have two X chromosomes, it doesn't mean that the genes on those chromosomes would all be the same, or even exactly half the same (factoring in the father's genetic makeup that he'll pass on to his offspring.)

Mutations happen all the time, so I'd say go with the mitochondrial DNA approach. Unless you plan on having the aunt and kiddo both have an extremely rare genetic disorder, in which case 'normal' DNA would show the same abnormal markers, and then probability, along with a few other gene sequences could determine the relationship from there.

MAP
07-03-2010, 09:08 AM
1. Like everyone else said, swab the cheek. Simple, harmless way to get a DNA sample.

2. Most likely the aunt and the child will show to be related.

The mother and the aunt will share some DNA as siblings, and even though the child will get only half of her DNA from the mother, it would be highly, highly unlikely that the child would have no DNA in common with her aunt. Thus, the child and the aunt's DNA will have more homology then the child would have with someone unrelated, and the results would say that there is a high probability that the two are related.

Does that make sense?


ETA
You don't need to go with mitochondrial DNA, nuclear DNA will work just fine. You need to look into how these paternity tests are done and how accurate these tests really are. Don't worry about mutations. DNA is big, and it is highly improbability that a mutation would occur in the sequences that are being examined. The sex chromosome (X) is only one of the 23 chromosomes that the child and aunt could share. It would be really, really weird if the DNA results did not suggest that the aunt and child are related. Let me put it this way, the child gets 23 chromosomes from her mother, how likely do you think it would be that none of those 23 chromosomes match the aunt's?

Project nachonaco
07-03-2010, 09:10 AM
1. Like everyone else said, swab the cheek. Simple, harmless way to get a DNA sample.

2. Most likely the aunt and the child will show to be related.

The mother and the aunt will share some DNA as siblings, and even though the child will get only half of her DNA from the mother, it would be highly, highly unlikely that the child would have no DNA in common with her aunt. Thus, the child and the aunt's DNA will have more homology then the child would have with someone unrelated, and the results would say that there is a high probability that the two are related.

Does that make sense?

I believe so. I've never seen a DNA database but I assume it breaks down into 'Possible Relatives' and 'Absolute Relatives' (or somesuch). And the aunt would show up in 'AR'.

Yes?

MAP
07-03-2010, 09:36 AM
I've never seen a DNA database either. But I would think that there would be something like that. Most likely dealing with a scale of probabilities of the two being related.

Although I doubt a scientist would use the term "absolute."

kelzey2
07-03-2010, 05:13 PM
When a DNA sample is taken, a DNA profile can be produced by locating certain repeated pieces of DNA.

http://i847.photobucket.com/albums/ab38/kelzey2/Capture.jpg

Of the two numbers under each heading, one is from the mother and one is from the father.

Not sure exactly how it would be done, but using that profile, a search could be done to find relatives. From what I know, I believe that the more closely related a person, the more numbers would match up. In other words, if a profile was found that had one number the same under every heading, the profile would be of a parent.

MarkEsq
07-03-2010, 05:34 PM
Is this a crime scene situation, or are they trying to locate a relative? I ask because if you are going for realism, there may be some issues with testing and comparing with a whole batch of employees. A police agency will probably not want to do that, honestly. But if it's just a matter of combing through the employees for the kid's relative, I have no comment. Other people know waaay more about DNA than I do, I just know about the police procedural angle. :)

PeterL
07-03-2010, 05:44 PM
That's not true at all. Familial traits will show up in both aunt and niece, they can track family trees back through many generations.

With strictly female lineage it's even easier because only women pass on the mitochondrial DNA.

see http://genealogy.about.com/cs/geneticgenealogy/a/dna_tests.htm

Furthermore, ALL humans share 99.9% of their DNA. Something with only 25% of a human's DNA wouldn't even be an animal.

It is true that only a limited set of genes have alleles; I thought that was understood. Of that set that have variations, It is perfectly possible that a neice (maternal) and her aunt could share only the X chromosome. Do the calculations. That is an extrme case, but it is completely possile. Similarly, it is possible for a mppair of brothers or sisters to shae no alleles except the sex chromosomes, and a brother and a sister could share no alleles; tht's very unlikely, but it is possible.

OneWriter
07-03-2010, 06:04 PM
Peter, you are misusing the word allele. All genes have alleles. They have two, in fact, because our chromosomes come in pairs.

To the OP: if you Google parental testing lots of companies that provide this service will come up. Call them or email them and ask them. Typically they analyze the Y chromosome for paternal testing and the mitochondrial DNA for maternal lineage. The answer is a probability because all they do is a statistical test. You do not get a Yes/No answer. You get a percentage. For DNA profiling I know they look at certain specific loci (typing the whole DNA is lengthy and darn expensive!), including a subset used for racial profiling, and there's a minimum number of loci that have to match in order to establish the identity. For parental testing that number I imagine would be lower.

Oh, and BTW: the loci they look at NEED NOT be on genes. Most of our DNA is NOT genes.

PeterL
07-03-2010, 06:22 PM
Peter, you are misusing the word allele. All genes have alleles. They have two, in fact, because our chromosomes come in pairs.

To the OP: if you Google parental testing lots of companies that provide this service will come up. Call them or email them and ask them. Typically they analyze the Y chromosome for paternal testing and the mitochondrial DNA for maternal lineage. The answer is a probability because all they do is a statistical test. You do not get a Yes/No answer. You get a percentage. For DNA profiling I know they look at a certain specific loci (typing the whole DNA is lengthy and darn expensive!) and there's a minimum number of loci that have to match in order to establish the identity. For parental testing that number I imagine would be lower.

You should study the subject.

Tsu Dho Nimh
07-04-2010, 12:08 AM
1. What's the easiest way to get a DNA sample from a two-year-old without making her upset?

Mouth swab or hair combings with intact roots.


2. When comparing her DNA, (let's say the place she was found at demands DNA samples from every employee), would her aunt show up in the list of DNA matches by approximation?

Depends ... if it's her mother's sister, there would be more matches than if you looked at an unrelated female but not as strong a match as if you tested her mother. If a perfect match (her identical twin) would be 100%, parents get a 50%, siblings get about 25% (bewcause of the way chromosomes assort) ... her mom or dad's siblings would be even less.

If you are looking through a DNA database for "potential family members" you have to set the search criteria accordingly and lower than if you are looking for parents.

The local cops caught a rapist who had never given a DNA sample ... enough of his relatives had, and they were getting consistent partial matches to one family group. They decided to see where the man had been at the times of the rapes, and he had been seen in the neighborhoods. that was enough to subpoena a mouth swab and there was the match they were looking for.

Tsu Dho Nimh
07-04-2010, 12:15 AM
I believe so. I've never seen a DNA database but I assume it breaks down into 'Possible Relatives' and 'Absolute Relatives' (or somesuch). And the aunt would show up in 'AR'.

Yes?

No. It's a bunch of numbers indicating the results of the DNA analysis. When you are looking for a match, you select the sample, tell the database (via a query form) to return all samples with numbers that match within XX%. Because some of the samples are from damaged DNA from crime scenes, you seldom ask for a 100% match, just the best ones.

A list comes back ... it's up to the humans running the search to decide who is and isn't related.

OneWriter
07-04-2010, 12:42 AM
No. It's a bunch of numbers indicating the results of the DNA analysis. When you are looking for a match, you select the sample, tell the database (via a query form) to return all samples with numbers that match within XX%. Because some of the samples are from damaged DNA from crime scenes, you seldom ask for a 100% match, just the best ones.

A list comes back ... it's up to the humans running the search to decide who is and isn't related.

Let me iterate that it's not a 100% match in the literal sense because as far as I know for parental testing they do not type the entire genome but loci that are highly variable across the population. It costs several thousand dollars to type the whole genome, and statistics teaches us that you really don't need to. When you get a minimum number of matches you enter the range of "high probable match" or in other words it is highly improbable that two individuals with so many similarities would not be related.

Even when they do type the entire genome, the machine inserts mistakes: what comes out of the machine are light reads that are then translated into integers. Depending on the technique used errors are more or less likely. Cloning also introduces errors, which is a step in the PCR process. That's why when you look at those diagrams they should always have a column with controls.

Hence, what you get out is a probability, and of course 99.99% is equivalent to a YES.

FrancisK7
07-04-2010, 03:57 AM
Hey there!

I'm a molecular biology major, so I hope I can be of some service.

1. A cheek swab for some epidermal cells is the easiest method to collect DNA
2. A DNA test will easily trace lineage using specific protein markers so yes, it will be more than easy to prove she and her aunt are related (even to which degree).

Hope this helps! If you have any more DNA-related questions send me a PM or e-mail !

Francis

PS. People have thrown numbers in this thread that are are either misleading or incorrect. We share 98.4% of our DNA with apes and probably 99.9% with other humans. The 0.1% difference accounts for all the physical differences. However, though we don't know exactly how much yet, we know most of a human's DNA is complete junk that serves no purpose whatsoever.

JulieHowe
07-04-2010, 04:36 AM
The cheek swabs inside the mouth are standard now. I used to work for an AIDS charity and we offered cheek-swab AIDS tests as early as 1997. This was a huge improvement over having to convince a client to give a blood sample, making the situation comfortable and non-threatening, and it was far safer for the person administering the test. The cheek swab method is also far better to convince would-be daddies to turn over a DNA sample, and it's less threatening than a court order from a family court judge demanding a blood sample.

I do recall a friend before 1997 being totally pissed because a blood sample was taken from her newborn infant by court order, after her ex-boyfriend decided to play stupid and say he wasn't the baby's father. A cheek swab DNA sample would have saved all three of them a lot of grief, but the test wasn't available in those days.

OneWriter
07-04-2010, 07:12 AM
However, though we don't know exactly how much yet, we know most of a human's DNA is complete junk that serves no purpose whatsoever.

I know you're all gonna roll your eyes at me, but sorry, I slightly disagree: we don't know what purpose it serves. I wouldn't say it serves no purpose whatsoever.

Back when we were looking at one gene at the time they found BRCA1 and BRCA2 and other markers for other types of cancers. Than they did the whole genome and lo and behold those disappeared and all statistical significance suddenly shifted to exactly those regions you call junk. So they may serve some purpose after all. One day we may even find out.

Maraxus
07-04-2010, 11:22 PM
Back up there, Tsu Dho said, siblings share about 25% of the markers. My "Gregor Mendel and his peas" level genetic knowledge tells me it should be 50% (25% for half-brothers, grandparents and aunt/uncles). Can someone confirm this?

OneWriter
07-05-2010, 12:46 AM
The 50% is an average; you get 50% of your DNA from mom and 50% from dad. However, these 50% are never the same as the chromosomes undergo a little reshuffling each time, hence the 50% you got from mom need not be the same 50% your sibling got from your mom. Same thing with the 50% you inherit from dad.

Now you pick any pair of siblings at random, they may share anything from 0% (if the random reshuffling gave each the exact complement of either 50%) to 100% (they got the exact same half from mom and the exact same half from dad). Of course these are the two extremes, and the average sits right in the middle at 50%.

That's why the 50% is an average. Any single pair of siblings may share either more or less than that.

FrancisK7
07-05-2010, 06:49 AM
I know you're all gonna roll your eyes at me, but sorry, I slightly disagree: we don't know what purpose it serves. I wouldn't say it serves no purpose whatsoever.

My statement was not an opinion. It is a scientific fact.

Enough breakthroughs have been made in genetic studies and our understanding of DNA in general is advanced enough to be able to confirm this. Like our appendix which serves no purpose whatsoever, it's an evolutionary byproduct. It was left there because it does no harm, but we know most of it is junk with absolute certainty.

It'd be overly technical to explain why, but any genetic textbooks written/updated in the past decade will be able to provide insight as to why we know this if you really want to get the details.

OneWriter
07-05-2010, 07:36 AM
The only reason why I object to the word "junk" is that if you take it out you no longer have viable DNA, meaning you don't get a live organism. So there's your function already: it's historically there, most of it inherited through the steps that made us turn from bacteria into human beings. Redundancy is one of Mother Nature's best strategies for survival. I guess I'm getting philosophical, but if nothing else there's your function already: cut it out of the DNA and you no longer have a human being. That's a long stretch from being junk. IMO.


It was left there because it does no harm, but we know most of it is junk with absolute certainty.


Darwin would certainly object to that. Darwin was a genius because he was the first to understand that the question "what purpose does it satisfy?" is a human question. Mother Nature doesn't work like that. We ask the question, we give the answer based on what we observe. We don't observe, we shrug and think, "oh well, it doesn't do any harm." That's the wrong thinking.

What you call junk is made of pseudogenes that have an evolutionary reason for being where they are. They are identical to genes we already have except they have frameshifts or stop codons that prevent them from coding. They are redundant. But like I said, throughout evolution redundancy has proven a great strategy for success.
I see now that even wikipedia calls them "junk DNA". Well that's really sad. I wanna think it's just a name, because truly, it's NOT junk in the literal sense. It's non-coding. Period. It does not mean we don't need it. If nothing else, it gives us a place in a phylogenetic tree.

FrancisK7
07-05-2010, 09:16 AM
When I meant "junk" it was the accurate and "scientific" term... some of our DNA is useless and achieves no roles. If you could easily splice out DNA out of a living being, you could just remove it and it wouldn't affect the person in any way. Like we only remove appendixes that cause trouble because in a healthy person you'd do more harm by putting him or her under the knife than just letting that useless organ be there.

For example, if our entire genetic material looked like this: (each row is a helix)

ABCDEFGHIJGKL
ABCDEFGHIJGKL

Though we don't know just how much, the amount of useful DNA that defines our "being" vs junk-space-filler DNA would be:

ABCD||EFGHIJGKL
ABCD||EFGHIJGKL

ABCD=useful part
EFGHIJGKL=junk, space filler... appendix dna (totally useless and carries nothing of importance to making you human)

We actually call it junk in the lab! ;)

OneWriter
07-05-2010, 05:05 PM
Oh, I see... you do the lab work, whereas I do the theory stuff. I look at phylogenetic trees and evolution and to me that stuff is important. It's like looking at the Rosetta stone. I see why you guys call it "junk" but like I said you shouldn't mean it literally. It doesn't serve a purpose in that one step you're looking at, which is when it splices it gets thrown away. Agreed. But there are mutations in there that distinguish one individual from another and that in some cases have been associated to cancer (not with absolute certainty, still...). I don't do full genome studies, but I remember a while ago, when they had just started it, a friend of mine told me that they would get significant associations exactly in those "junk" regions. That's statistics of course, and as we all know, "there's lies, there's damn lies, and then there's statistics," but still...
Sorry, I didn't mean to be pedantic, I just get a little iffy when I see the words "absolutely" and "certainty" applied to science.... That's all.