Gedmatch rare SNP search

[linkus]

Alright then, my country list, in descending order by number of times observed:

linkus
4	Guatemala 
3	Sicilian
3	County Kildare, Ireland 
3	Ashkenazi
3	Caucasus 
3	Mexico 
2	Abruzzo/Abruzzi, Italy 
2	Denmark 
2	Italy
1	Britain
1	Donegal, Ireland
1	Germany
1	Greece
1	Lancashire, England
1	Mayan 
1	North East Spain
1	Puerto Rico
1	Quebec 
1	Southern Italy
1	Spain


mother
5	Guatemala 
4	Denmark 
2	Ireland
2	Spain
1	Britain
1	Caucasus 
1	County Kildare, Ireland 
1	German Jew 
1	Germany
1	Greece
1	Lancashire, England
1	Mayan
1	Mexico 
1	North East Spain
1	Northern Portugal 
1	Puerto Rico
1	Sicilian
1	Southern Italy


maternal grandfather
5	Guatemala 
3	Colombia 
3	Mexico 
2	Ashkenazi 
2	County Kildare, Ireland 
2	Mayan
2	North East Spain
2	Ontario
1	Caucasus 
1	Germany
1	Lancashire, England 
1	Northern Portugal 
1	Puerto Rico
1	Sicilian
1	Spain


paternal grandfather
4	Spain
4	Ashkenazi 
2	Netherlands 
2	Abruzzese
2	Eritrea 
1	Britain
1	Caucasus 
1	County Kildare, Ireland 
1	Germany
1	Ireland
1	Italy
1	Lancashire, England
1	North East Spain 
1	Quebec 
1	Sicilian


paternal grandmother
10	Mexico
5	Guatemala 
5	Ashkenazi
4	Mayan
2	Caucasus 
2	Colombia 
2	County Kildare, Ireland
2	Denmark 
2	Sicilian
1	Abruzzo/Abruzzi, Italy 
1	Donegal, Ireland
1	Greece
1	Italy
1	Lancashire, England
1	Netherlands
1	Spain

[Vadim Verenich]

The same mutations could easily have independently occured in seperate places

The same mutations could have theoretically emerged in unrelated populations, but only few of them got fixated due to the natural selection. Like that geezer Darwin who used to say:" the survival of the fittest shapes the path of evolution".

[Jonny]

The same mutations could have theoretically emerged in unrelated populations, but only few of them got fixated due to the natural selection. Like that geezer Darwin who used to say:" the survival of the fittest shapes the path of evolution".

But we are talking about rare mutations here, not common ones. Ergo they haven't yet been that successful (whatever that means) in an evolutionary sense.
Also they may not, and probably don't, have any positive or negative effect on survival and fitness.

[Vadim Verenich]

But we are talking about rare mutations here, not common ones. Ergo they haven't yet been that successful (whatever that means) in an evolutionary sense.
Also they may not have any positive or negative effect on survival and fitness.

If mutation doesn't have any positive or negative effect, then it is a neutral mutation.
However, the rare mutations harbored by mankind population-wise are usually (by)product of the selection.

[Mihajlo]

My mother's Rare snp graph vs her BGA.








rare snp on chromosome 1, on asian segment.


rs2000161 1 206740974 A G 0.7436 Ashkenazi

frequency : http://www.ncbi.nlm.nih.gov/SNP/snp_...0161#Diversity

rs17012595 1 206707872 C T 0.6874 SYT14 Ashkenazi

frequency : http://www.ncbi.nlm.nih.gov/SNP/snp_...2595#Diversity


rs5015728 1 218027676 C T 0.4588 Colombia Guatemala Mayan

frequency : http://www.ncbi.nlm.nih.gov/SNP/snp_...5728#Diversity


rare snp on chromosome 6 on mideast segment.


peaks in japenese

rs11751545 6 32149021 A C 0.4477 TNXB Donegal, Ireland
frequency : http://www.ncbi.nlm.nih.gov/SNP/snp_...1545#Diversity


peaks in japenese

rs11754464 6 31831714 C T 0.4936 MSH5 Donegal, Ireland
frequency : http://www.ncbi.nlm.nih.gov/SNP/snp_...4464#Diversity


peaks in japenese


rs11757034 6 32114701 C T 0.4135 CYP21A2 Donegal, Ireland
frequency : http://www.ncbi.nlm.nih.gov/SNP/snp_...?rs=rs11757034




Donegal, Ireland while it peaks in japenese population according to ncbi

[Vadim Verenich]

That's not a problem at all.
The REAL problem is that many alleles in Rare SNP search are misreported.
The NCBI SNP database reports alleles on the forward strand of DNA, while Gedmatch Rare SNP search (in some of rare snps) rare alleles on the reverse strand. Such a situation will inevitably cause a lot of confusion.

I'll provide some examples of discrepancy between NCBI and Gedmatch later.

[Jonny]

If mutation doesn't have any positive or negative effect, then it is a neutral mutation.
However, the rare mutations harbored by mankind population-wise are usually (by)product of the selection.

Seems like they are mostly rare because they are neutral.

They are not negative, which means they are not selected against until they disappear.
They are not positive, so they don't spread to the wider population.

[Vadim Verenich]

Seems like they are mostly rare because they are neutral.

They are not negative, which means they are not selected against until they disappear.
They are not positive, so they don't spread to the wider population.

Ok, it'll take a lot off time to explain the modern theory of natural selection from the viewpoint of molecular biology.

Due to the limited schedule, i'd rather focus on more relevant examples.

But if you are indeed interested in the topic, i'd advise you to read "Principles of Population Genetics", Fourth Edition by Daniel L. Hartl and Andrew G. Clark. This is a really good reading.

---------- Post added 2012-07-10 at 19:13 ----------

Now let's turn back to our SNPS for good.
As i said earlier, i'd like to illustrate the most obvious cases of discrepancies between the Gedmatch results and SNP reports from NCBI dBSNP database.

(note that the examples below are taken from the same Gedmatch Rare SNP query)

Let's look at rs10064056 and compare its alleles against the alleles of the same snp reported in NCBI.

Gedmatch: rs10064056 5 15473775 C T 0.3636
NCBI: rs10064056 RefSNP Alleles: C/T(ancestral allele is T) forward strand +

It looks like that those alleles reported by Gedmatch are matching the refSNP alleles of rs10064056.

Unfortunately, this is not always a case.

Consider another reported 'rare' SNP:

Gedmatch results: rs5744265 11 111524133 C T 0.8317 IL18
NCBI Reference SNP(refSNP) Cluster Report: rs5744265 RefSNP Alleles: A/G

From what i can see in Gedmatch results for this particular SNP, the alleles of SNP are being reported in a 'flipped' form (i.e., alleles on the reverse strand of DNA -).

I was expecting something different. Running the rare SNP search on the SNP set one half of which belongs to the forward strand of DNA and another belongs to the reverse one, doesn't make much sense

[Jonny]

Ok, it'll take a lot off time to explain the modern theory of natural selection from the viewpoint of molecular biology.

You can always tell a Yorkshireman, but you can't tell him much.

[Vadim Verenich]

But we are talking about rare mutations here, not common ones. Ergo they haven't yet been that successful (whatever that means) in an evolutionary sense.
Also they may not, and probably don't, have any positive or negative effect on survival and fitness.

Ok, of course we can ask ourselves whether these reported SNPs are truly rare mutations or rather statistical fluctuations.
At the same time, some of the reported rare snps are in linkage with very well-known gene domains (check the 7th column of your report). When SNP is in the linkage, there is no doubt that it is affected by selection/good genetic background. If this SNP was a result of neutral mutation, the block of linkage disequilibrium would have been reshuffled by recombination and this mutation gone long time ago.