PDA

View Full Version : Help to understand the prevalence of false positives please



DillonResearcher
03-20-2017, 08:13 PM
I've recently been looking over some Geno results for other testers positive for R-L135 (two SNPs below L2) and would like to understand a bit more about Geno testing. From what I understand it's not Next Gen Sequencing of course but isn't Sanger sequencing either and is some other sort of chip based SNP test, I think?

What I'm really wanting to better understand is the issue of false positive results, I've heard people mention the Geno test as giving false positive results sometimes but want to know how common this is and if there is any way to spot false positive results?

RobertCasey
03-21-2017, 05:06 PM
I am not sure what you mean by GENO testing. NatGeo testing is product sold by National Geographic using very inexpensive chip array testing equipment. This technology has a hard time reading certain area of the Y chromosome but these bad reads are quickly replaced with no reads since results are not consistent. Since this is a static test (no variability) and is an economical test, it only reports around 23,000 YSNPs.

NGS testing uses another technology is primarily used for discovering new private YSNPs that will become tomorrow's new branches. It also reports around 65,000 known YSNP mutations. It usually reports five to ten new private YSNPs but we just received our second Y Full Elite 2.1 test and received no new private YSNPs, so YSNP is reaching its current limits in some areas that are well tested.

Sanger Sequencing is yet another technology used for testing of individual YSNPs. Also, YSEQ uses Sanger Sequencing in their tiered approach for SNP packs as well. YSNPs being tested by Sanger Sequencing have trouble with testing YSNPs found in unstable areas of the Y chromosome that NGS tests can read.

NGS testing results is extremely variable for YSNP positions, varies between testers and has major variations between different NGS tests and WGS tests. Due to smaller read lengths of the existing economical NGS tests, some YSNPs are harder to read than others, so some YSNPs are questionable calls, no-calls or good quality calls. For some YSNPs, there is inconsistent reporting results but these YSNPs are not extremely common but do happen a lot. Various NGS tests (Big Y vs. YElite) have significant differences as well. To reduce testing costs, FTDNA only reads YSNP rich areas and its read length is only 150 base pairs vs. 250 from Full Genomes. YElite 2.1 can read 30 % more YSNPs than Big Y. FGC WGS Long read can read twice as many base pairs of Big Y since its read length is 1,000,000 base pairs. But the price of $2,950 restricts its current usage to only a few testers.

DillonResearcher
03-21-2017, 05:18 PM
I am not sure what you mean by GENO testing. NatGeo testing is product sold by National Geographic using very inexpensive chip array testing equipment. This technology has a hard time reading certain area of the Y chromosome but these bad reads are quickly replaced with no reads since results are not consistent. Since this is a static test (no variability) and is an economical test, it only reports around 23,000 YSNPs.

NGS testing uses another technology is primarily used for discovering new private YSNPs that will become tomorrow's new branches. It also reports around 65,000 known YSNP mutations. It usually reports five to ten new private YSNPs but we just received our second Y Full Elite 2.1 test and received no new private YSNPs, so YSNP is reaching its current limits in some areas that are well tested.

Sanger Sequencing is yet another technology used for testing of individual YSNPs. Also, YSEQ uses Sanger Sequencing in their tiered approach for SNP packs as well. YSNPs being tested by Sanger Sequencing have trouble with testing YSNPs found in unstable areas of the Y chromosome that NGS tests can read.

NGS testing results is extremely variable for YSNP positions, varies between testers and has major variations between different NGS tests and WGS tests. Due to smaller read lengths of the existing economical NGS tests, some YSNPs are harder to read than others, so some YSNPs are questionable calls, no-calls or good quality calls. For some YSNPs, there is inconsistent reporting results but these YSNPs are not extremely common but do happen a lot. Various NGS tests (Big Y vs. YElite) have significant differences as well. To reduce testing costs, FTDNA only reads YSNP rich areas and its read length is only 150 base pairs vs. 250 from Full Genomes. YElite 2.1 can read 30 % more YSNPs than Big Y. FGC WGS Long read can read twice as many base pairs of Big Y since its read length is 1,000,000 base pairs. But the price of $2,950 restricts its current usage to only a few testers.

Thanks for the response, it's the NatGeo test that I was referring to. The reason I was asking is that I've been going through all of the Genographic Project's R-L135 data and was interested to see that there were several L135+ people with Irish paternal roots which was a pleasant surprise as I thought that I was the only L135+ person so far with Irish ancestry. I knew that there were problems with some NatGeo results and so just wanted to know how likely it is/how common for the result to be a false positive.

I'm trying to find out more about L135 in general and 8 have tested L135+ with FTDNA but around 70 tested L135+ with the NatGeo test and so with the majority of the information regarding distribution coming from NatGeo testing I was interested to learn a bit more about it.