PDA

View Full Version : YFULL compared to FTDNA : 2 new mutations



palamede
12-11-2014, 04:14 PM
In August 2010, I ordered the mtfullsequence of FTDNA. I was found perfect rCRS with no mutation from rCRS; therefore H2a2a1.

This year, Yfull analysed my Big Y BAM and they found I am H2a2a1 really but with 2 mutations from rCRS : C3460T and C6277T.

I am troubled, I think Yfull is correct and in this case the FTDNA list of matches is not correct.

Foy you, I should believe Yfull or FTDNA. Do you know similar disgreement ?

jesus
12-11-2014, 04:16 PM
This might help: http://dna.jameslick.com/mthap/

Petr
12-11-2014, 04:27 PM
mtDNA markers obtained from BigY BAM file analysis are very unreliable. I got many false mutations.

VinceT
12-12-2014, 06:44 AM
If you want a second opinion, Full Genomes Corp (http://napobo3.lk.net/Procedure_for_ordering_and_registering_a_3rd_party _data_analysis.html). has a coupon code to reduce their BAM analysis cost from $40 to $25 per sample: IP109. I presume it is still valid.

I would still tend to trust FTDNA's report for mtDNA over YFull's; a perfect rCRS match is a bit surprising but it does exist, and the mtDNA data in BAM files from both BigY and Full Genomes can be error prone due to alignment issues and poor coverage. Conversely, I have known FTDNA to screw up from time to time as well.

palamede
12-12-2014, 03:36 PM
In May, I ordered the FGC analysis of my BIG Y. I received a mtdna report of the variants with respect to Yoruba reference sequence.

I didn't examine the file because a lot of ? and ??.
Today I am looking at with more care by comparing my values from AF347015 (Yoruba reference) in PhyloTree. For the bigger part of the ADN, the addresses of Yoruba reference have a difference of , -1 or -2 versus the Phylotree addresses. In general from L3e2b1a1 (AF347015) to H2a2a1 , I found the addressess with following differences (0, -1,-2) according to the part of the ADN mitochondrial (often with the real substitute nucleotide but sometimes with ??).

Otherwise it remains
- some part if the adn have ? for the 2 nuleotide (the substitued and the substitude).
- some new SNP (or indel) adresses with values of the 2 nucleotides or 1 subtituted nucleotide and ?? for the substitude one. Very troubling because I don't know if real positive SNPs or false positives.
- For the 2 further SNPs of Yfull, I found them with differences in the addresses of their portions of the ADN :

______YFULL_______________________FGC
______G3460T___________________G3461G/T
______G6278T___________________G6277??

I am more and more perplexed. I can be sure to be H2a2a1, but with no further mutation or 2 or more mutations ?

VinceT
12-13-2014, 07:16 AM
palamede,

Thanks. I took a look at the chrM alignments in a few Big-Y BAM files I have on hand, and to be honest, I don't find them extremely convincing. It appears the alignment parameters were treated a bit too loosely, and read fragments that out to belong elsewhere are mis-mapped to the mtDNA sequence alignment, reinforcing my suspicion that YFULL are reporting false positives or false negatives as the case may be. Unfortunately in this case, Full Genomes wouldn't be able to do much better without extracting the raw FASTQ reads and regenerating the alignment. It's the old "garbage in, garbage out" scenario.

Big Y was not originally intended for robust mtDNA analysis. Although FGC claims to have improved their mtDNA analysis pipeline a few months ago, I do not know if they implemented re-mapping of mis-aligned reads.

Vince T.