PDA

View Full Version : atDNA run at ancestry or FTDNA. HOw does it handle heterogenous SNPs



Jan_Noack
02-06-2018, 09:23 AM
Hi everyone,

I was wondering how the atDNA tests handle heterogeneous SNPs. I have tested both at ancestry and FTDNA n a few kits but not looked in detail at results. I have looked at matches though.

Is only one SNP reading recorded for each homologous SNP on the pair of chromosomes .e.g. there are 2 chromo 1's.. on the same SNP one side of the chromosome reads C and the other side reads G Are both recorded if different? ie heterogeneous? Or is only one result recorded (maybe the most common one found on the no of reads)? If only homologous results are recorded, or if only one (majority of reads) result is recorded , then (except for the Y and the X for the male passed on) maybe the parent may not match the child as far as test resuts go on all expected SNPs. Obviously both parents each pass on one half of the pair in the child. The mother's half on say chromo1 is a unique combo of her parents recombined strands for chromo1 , one from each maternal grandparent that the mother carries as her pair of chromo1s. (Similarly with fathers passed on chromo1 from paternal grandparents).
The passed on SNPs in the child from the mother will be on the one of the mother's two strands..but will these SNPs all be picked up by the atDNA testing of the mother (ie is it heterogenous or does it record both SNP at each homologous position?..or only the one "most" tested value). If notbthe mother mayhave a different SNP recorded than the child for a given position as it was the other SNP tat was passed on to the child (from the mother's pair)?

Luckily SNPs are not passed on individually but in large bundles ..as genes/ genetic traits etc.like your looks, colouring, and down thru the generations and also the recombination is in no way random but depends on many factors such as dominance and combos of Chromos crossing together.. making matching way easier to see when larger groups pass down together.

My query is IF a group is passed down from say the mother to the daughter will the atDNA test always pick this up..or does the test only record the most tested SNP or , does the atDNA matching program only use the most frequently occurring SNP tested for comparison..and may possibly "miss" a larger block altogether (as it recorded the other side on the chromosome pair only, of say the mother , as this was the bit that got the (larger no of) reads?. Hope this makes sense as I am finding this difficult to put in writing and if it makes no sense everyone, I think I'll just quietly give up. Thanks for reading and trying to understand.

geebee
02-06-2018, 11:45 PM
Hi everyone,

I was wondering how the atDNA tests handle heterogeneous SNPs. I have tested both at ancestry and FTDNA n a few kits but not looked in detail at results. I have looked at matches though.

Is only one SNP reading recorded for each homologous SNP on the pair of chromosomes .e.g. there are 2 chromo 1's.. on the same SNP one side of the chromosome reads C and the other side reads G Are both recorded if different? ie heterogeneous? Or is only one result recorded (maybe the most common one found on the no of reads)? If only homologous results are recorded, or if only one (majority of reads) result is recorded , then (except for the Y and the X for the male passed on) maybe the parent may not match the child as far as test resuts go on all expected SNPs. Obviously both parents each pass on one half of the pair in the child. The mother's half on say chromo1 is a unique combo of her parents recombined strands for chromo1 , one from each maternal grandparent that the mother carries as her pair of chromo1s. (Similarly with fathers passed on chromo1 from paternal grandparents).

Two values will be reported at each genomic position tested. They can be the same -- for example, AA. Or they can be different -- perhaps AC or AG.


The passed on SNPs in the child from the mother will be on the one of the mother's two strands..but will these SNPs all be picked up by the atDNA testing of the mother (ie is it heterogenous or does it record both SNP at each homologous position?..or only the one "most" tested value). If notbthe mother mayhave a different SNP recorded than the child for a given position as it was the other SNP tat was passed on to the child (from the mother's pair)?

With the exception of a miscall or no call, at least one letter out of every SNP pair in a parent should match at least one letter out of every SNP pair in a child. So if the child has, say, AA CT GG AG GT, for those same positions the parent might have AT CC GT AA GT. Only one has to match in each pair, though both might match.


Luckily SNPs are not passed on individually but in large bundles ..as genes/ genetic traits etc.like your looks, colouring, and down thru the generations and also the recombination is in no way random but depends on many factors such as dominance and combos of Chromos crossing together.. making matching way easier to see when larger groups pass down together.

Pretty much, although the unit that is passed on is not even the gene, but entire chromosomes. For each of the 23 chromosomes passed on from parent to child, there are actually three possibilities. The chromosome can contain DNA from only that parent's mother; from only that parent's father; or from both of the parent's parents. For example, my daughter's paternal copy of chromosome 9 is indistinguishable from the copy I inherited from my father -- that is, it contains no DNA from my mother. On the other hand, my daughter's paternal chromosome 7 contains only DNA from my mother (none from my father). Most of the chromosomes she inherited from me, however, include DNA from both of my parents.


My query is IF a group is passed down from say the mother to the daughter will the atDNA test always pick this up..or does the test only record the most tested SNP or , does the atDNA matching program only use the most frequently occurring SNP tested for comparison..and may possibly "miss" a larger block altogether (as it recorded the other side on the chromosome pair only, of say the mother , as this was the bit that got the (larger no of) reads?. Hope this makes sense as I am finding this difficult to put in writing and if it makes no sense everyone, I think I'll just quietly give up. Thanks for reading and trying to understand.

Again, except for a few miscalls or no calls, the test should show that parents and children have at least one SNP in common at every genomic position tested. For example, at rs8176719 (located on chromosome 9), I have DI, my daughter has DI, and my wife has DD. The "D" in this case is used to mean that the original SNP, "G", has been deleted. Some companies will report this SNP as -/-; -/G; or G;G instead.

If you only looked at the values for my daughter and me, it would be impossible to tell whether she inherited the "D" (-) or the "I" ("G") from me. But since the only possibility for her inheritance from her mother here is "D" ("-"). then it's obvious that the "I" ("G") is what came from me.

My father tested at 23andMe rather than Ancestry, but he has "II" at this SNP. So I inherited the "I" from him, and that's what I passed on to my daughter -- along with all the rest of chromosome 9.

geebee
02-07-2018, 12:16 AM
Incidentally, you can easily download your Ancestry data file to see what it actually looks like. From "Your DNA Results Summary", click on "Settings" in the upper right corner. You should then see a page of "test settings", including a button for requesting to download your DNA file.

After you follow the instructions to download your file, you can open it and take a look at the SNPs identified by the test. At this point, there are possibilities such as uploading to either Promethease or GEDmatch to learn even more. And in the case of GEDmatch, you can compare your results even to those who tested at another company, as long as they've uploaded their DNA file to GEDmatch. There's also a chromosome browser available -- something Ancestry steadfastly refuses to provide.

Why is a chromosome browser important? Well, for one thing, just because you may have a match with two individuals who also match each other, does not mean the three of you have common ancestry. But, if any matches overlap -- provided they're on the same side, of course -- then common ancestry is likely for all three.

There are quite a few other tools available at GEDmatch, too, but all of them depend on your obtaining your DNA file and uploading it.