MISCELLANEOUS NOTES ON DNA MATCHES AND GENETIC GENEALOGY
Miscellaneous notes from a variety of sources relating to DNA matches and genetic genealogy in general. I may eventually get around to having individual webpages for some of these, but at the moment this is just a place for me to dump useful information!
Most Recent Common Ancestor (MRCA)
- According to Tim Janzen on Rootsweb GENEALOGY-DNA Mailing List (Here), which is a little disillusioning (with respect to atDNA):
"Just because you share a 16.9 cM HIR (half-identical region) with someone don't assume that you are related to that person within 5 generations. The GEDmatch suggested generations to MRCA should be taken with a grain of salt. As a general rule, I assume that any match who shares 50 cMs or less with one of my relatives or me could be as distantly related as 10th or 20th cousin."
However, Tim also comments in another post (Here) that compared with now (in 2013) when we are "Pioneers" in the field and the total number of people tested for atDNA is relatively low (probably 500,000-550,000) but that in 10 years time there will be several millions and there will "be more relatives for sure". He cautions about waiting for 10 years before testing, however, because "a number of the accounts will be abandoned by then and there won't be anyone able to share the pedigree chart for those accounts if they didn't make a pedigree chart available before the person died or otherwise abandoned the account."
So I recommend that you don't delay being tested, as you will be WAY behind those of us who have already hopped onto this bandwagon.
- Some companies have reported on HVR3 for positions 391 to 590 but Family Tree DNA does not use HVR3. At FTDNA, HVR2 includes positions 1 to 579. HVR1 includes positions 16001 to 16569. So therefore you need to ignore HVR3. Exclude your coding region (CR) when checking for matches at Mitosearch and SMGF.org. Make sure you enter rCRS values at these sites and not the default RSRS values that have become standard only recently. FTDNA gives you a choice between these two options. Source: Kathy Johnston posting – see also HVR3 vs Coding Region.
Sound advice from Jim Bartlett, DNA-NEWBIE mailing list Message 36275 (30 Nov 2013) [my additions in square brackets]:
The folks I know about who have been Chromosome Mapping for a while, set a threshold of 7 cM (maybe 5 cM for known close relatives) for the segments we use. We do this for several reasons:
- Every Match has at least one 7 cM segment. So they are all still in the mix
- Most Matches are true cousins on the 7 cM segment.
- This keeps our spreadsheet to a manageable size.
- Smaller segments are far more likely to be IBS [Identical by State – that is, by coincidence] and then introduce wild goose chases into our mapping.
- The Triangulation process should be based on the same rules, for these same reasons. Overlap 7 cM or more.
- You'll notice FTDNA, 23andMe and GEDmatch all recommend a threshold – this is based on experience with lots of results.
- It's hard enough to map with larger segments, it gets very complex when errors are introduced to the mix.
- There is no guarantee that all 7 cM segments are IBD [Identical by Descent – that is, inherited] either, just a larger share of them. Some people set their minimum thresholds at 10, 12 or 15 cM for these reasons. (Personally, I don't like to leave identified Matches out of the mix.)
- I can't say all 2-4 cM segments are IBS – some will be IBD, but most will be IBS and they'll give you false leads. This wastes your time and your Matches' time; and with Triangulation, it could well pass this flakey info on to others.
Source: Tim Janzen from Rootsweb AUTOSOMAL-DNA mailing list, 12 Apr 2014. My extraction and interpretation from this:
- When using 23andMe data (autosomes + X-chr), divide the total cM by 74.4 for females (two X-chromosomes) and 72.6 for males (one X-chromosome) to get the percentage of the genome that two people share. In Build 37, per Family Inheritance: Advanced, there are 7438.6 cM when combining the autosomal DNA and the two X chromosomes in females and 7256.8 cM in males – and 7074.6 cM for just the autosomes.
- When using Family Finder (FTDNA) data (autosomes + X-chr), divide the total cM by 71.6 for females (two X-chromosomes) and 69.6 for males (one X-chromosome) to get the percentage of the genome that two people share, or if just using the autosomes, divide by 67.7 (both sexes). There are 7158.06 cM when combining the autosomal DNA and the two X chromosomes in females and 6962.13 cM in males – and 6766.2 cM for just the autosomes.
- When using 23andMe data divide the total cMs by 75 to get the percentage of the genome that two people share; there are 7494.8 cM when combining the autosomal DNA and the X chromosome per Family Inheritance: Advanced
When using FTDNA’s Family Finder divide the total cMs by 68 to get the percentage of the genome that two people share.
The FTDNA number is lower than the 23andMe number since FTDNA doesn’t include the X chromosome in the calculations
Thresholds for Matching
- Thresholds for matching (Source: Tim Janzen from Slides 22 and 25 of PPT):
23andMe: 7 cM and 700 SNPs for the first segment; additional segments can be as short as 5 cMs as long as they contain 700 SNPs (and when already sharing genomes, only the latter criterion is used) – note below for X-chr thresholds
FTDNA: 7.7 cM and 500 SNPs for the first segment [I think I've heard somewhere (?CeCeMoore or Shannon Christmas) that, in addition, the total combined cM from all segments down to 1 cM must be at least 20 cM]
AncestryDNA: 5 million base pairs for the first segment, which should on average include 1229 SNPs – based on a report from Tim Janzen after the 2014 Rootstech Conference (Feb 2014), senior people from AncestryDNA indicated that for the past few months that matching has switched from using 5 Mb pairs to 5 cM
How much of Parents' DNA do groups of siblings share?
- At any location, there is 1/2 (50%) chance that the segment will not be inherited by a child and with n siblings, there is 1/2^n chance that it will not be inherited by at least one of the n siblings.
- So 1-(1/2^n) of a parent's genome coverage will be inherited by the n siblings, i.e.:
- One child inherits exactly 50% of each parents' genome
- With 2 siblings, on average 75% of the parents' genomes will have been inherited overall
- With 3 siblings, on average 87.5% of the parents' genomes will have been inherited overall
- With 4 siblings, on average 93.75% of the parents' genomes will have been inherited overall
- With 5 siblings, on average 96.975% of the parents' genomes will have been inherited overall
- An interesting message string on DNA-Newbies board: 50136. There are some different views for the likelihood of a group of 3rd/4th/5th, etc. cousins will share the same DNA segment. I don't fully understand the calculations, but I'm with David Watson (19 Oct 2015 message) on this – he provides the Excel formula and a sample spreadsheet.
Criteria on AncestryDNA for Estimates of Relationships
According to Tim Janzen (originally reported by Angie Bush) and reported on the Rootsweb GENEALOGY-DNA Mailing List (see Here), Kenny Freestone of Ancestry.com included the following information in a presentation in May 2014 as the categories and criteria for various relationships on AncestryDNA:
- Parent/child (parent, child – immediate family member)
- Immediate Family (immediate family to close family)
- Close Family (close family to 1st cousin)
- Second cousin (1st to 2nd): 200,000,000 base pairs
- Third cousin (3rd to 4th): 150,000,000 base pairs
- Fourth cousin (4th to 6th): 100,000,000 base pairs
- Distant cousin (5th to 8th – moderate confidence): 30,000,000 base pairs
- Distant cousin (5th to 8th – low confidence): 20,000,000 base pairs
- Distant cousin (5th to 8th – very low confidence): 5-10,000,000 base pairs
Also, the switch from 5 million base pairs as the criteria for a match to 5 cM occurred ~Jan 2014.
Obtaining "exact" segment boundaries on 23andMe: Source: Ann Turner (providing instructions from Tim Janzen), Genealogy-DNA #1390054498, 18 Jan 2014:
- Log onto your 23andMe account using Firefox as your browser.
- Go to the Family Inheritance feature at https://www.23andme.com/you/inheritance and then go to "Genome View".
- In the Firefox menu bar click on "Tools", then click on "Web Developer", then click on "Web Console".
- Right click anywhere in the console. This should bring up a menu that has 3 options in it. Click on "Log request and response bodies".
- Select the two people you want to compare using the dropdown menus in Family Inheritance and wait for the matching segment information to appear.
- Go to the Web Console and scroll down to the very bottom of the console where you should see a line that is the 2nd row from the bottom that reads: "POST https://www.23andme.com/you/ibdview/" as well as an HTTP extension after that. Click on this line.
- This brings up a box that says "Inspect Network Request". Scroll down to the very bottom of the box to the "Response Body" section. In this section you will see a subsection called "intervals". This subsection contains the start and stop positions for each matching segment and is organized by chromosome.
- Copy the start and stop positions for each matching segment into another spreadsheet so that you can save it.
Why generation estimates based on segment length cannot be accurate,
Source: Jim Bartlett, Genealogy-DNA #1390023058, 18 Jan 2014:
You do get exactly half of your atDNA from each parent. And each parent does this by giving you exactly half of their own atDNA. Each parent has two sets of chromosomes 1-22; they each give a child one set of chromosomes 1-22; and then the child has two sets of 1-22 - just like every human. The issue is that each chromosome is made up of random segments from their ancestors - at many levels. And each chromosome that is passed to a child is made up of big parts from the two chromosomes the parent got from their parents. This process is recombination and at each generation a few segments from the preceding generation are subdivided (usually from 0 to 3 subdivisions for each chromosome for each generation) Since there are relatively few subdivisions (crossover points) in each generation, most of the smaller segments (but still over 7cM) are not subdivided. So many pass from parent to child for a generation (or few) about the same size.
I used to call these sticky segments, but I now realize they are the result of a natural process and are actually to be expected.
Hope this helps in understanding why generation estimates based on segment length cannot be accurate.
The Significance of Segment Overlaps
A good discussion on the Rootsweb Autosomal-DNA Mailing List, April 2014, between Don Worth (1396551256 and 1396586357), Jim Bartlett (1396584058 and 1396634930), and Tim Janzen (1396597276).
Tips for Setting up Spreadsheets for Triangulated Groups/Chromosome Mapping for Multiple Siblings and Phasing when Parents' Data Not Available
Source: String of Messages on Rootsweb Autosomal-DNA Mailing List, April 2014, from Jim Barlett (1397627125) and Tim Janzen (#1397623742)
For triangulated groups, overlapping segments, and chromosome mapping, their advice is to make separate files for each relative (but I think this could be a separate worksheet within the same workbook file), making yourself the priority file. If you and a sibling share the exact same HIR (half-identical region), enter the data only in your file; but if someone matches a sibling but not you, enter the data in the other sibling's file. And if multiple siblings, designate a 2nd priority, then a 3rd priority sibling and continue in this same manner. Tim's message also covers phasing using data from children and cousins.
Where are the Centromeres Located on Each Autosomal Chromosome?
See FTDNA FAQ Here and ISOGG page on the Centromere (which gives slightly different locations).
Which DNA strand is reported?
Positive strand (= forward, coding, nontemplate, sense, or plus strand) which is read in the direction from 5' to 3' direction. The opposite strand is referred to as the negative, noncoding, template, antisense, or minus strand and is read from the 3' to 5' direction.
Showing Hidden Family Finder Matches on FTDNA
Thanks to Deborah Castillo for posting this on the GEDmatch User's Group on 20 Jul 2016 (Permalink available for Group Members), to show "hidden" names. Only surnames are shown (sometimes with Ms./Mr./Dr., but no first names), but this could still be very helpful for those looking for biological family. To see these:
- Click on MyFTDNA.
- Select Family Finder.
- Check Advanced Matches.
- Check the Family Finder box.
- Check Show only people I match in all selected tests.
- Choose Results Per Page 500 [or fewer can be used].
- Press Run Report.
- When results are there, go to the far right, and press Family Finder once to re-sort. Every name in black on the far left that cannot be clicked on is a hidden kit [only surnames are shown].
Segment Length Detection Thresholds: DNA Relatives only identifies people who shares a DNA segment of at least 7 cM and 700 SNPs long
How can a child have a match that doesn't appear in either of their parents' DNA Relatives? It does happen that a child sometimes shows a match that's not visible in either parent. In these cases, it's almost always the case that the match is indeed present in one of the parents, but the segment identification algorithm has missed the segment in the parent. This typically happens with short segments, where it is more likely that a candidate segment will fall just under one of the algorithm's thresholds.
Why do I have some matches listed in Countries of Ancestry which aren't visible in DNA Relatives? Countries of Ancestry shows all matches down to 5 cM segments. In contrast, DNA Relatives uses a threshold of 7 cM segments to display matches. Because the detection threshold is lower, more matches can be shown in Countries of Ancestry.
- Best Way to Download Your 23andMe Ancestry Composition Data (The Rose Bush, Kasandra Rose, 1 Feb 2013)
- Differences between Version 3 vs. Version 4 chip at 23andMe (the transition occurred Dec 2013-Jan 2014), indicating that the older V3 chip was better than the current V4 chip for the number of SNPs tested for autosomal and X-chromosomes:
# autosomal SNPs tested: 930,381 (V3) vs. 577,382 (V4)
# X-chromosome SNPs tested: 26,007 (V3) vs. 19,487 (V4)
# Y-chromosome SNPs tested: 1,766 (V3) vs. 2,329 (V4)
# mt-DNA DNPs tested: 2,737 (V3) vs. 3,154 (V4)
Source: Slides 27 and 28 of Tim Janzen's presentation (Using Autosomal DNA Testing as a Means to Trace Your Family Tree) to Southern California Genealogy Society (4 Jun 2015) available at http://tinyurl.com/ntbe3zs
cM Calculation Tool: (Source: Ann Turner, DNA-Newbie Message 38320)
- If you match one person with the segment boundaries at 10000000 and 50000000 and another person with the segment boundaries 30000000 and 70000000, the boundaries for the overlap are 30000000 and 50000000.
- Then use http://compgen.rutgers.edu/mapinterpolator (Build 36) http://compgen4.rutgers.edu/RutgersMap/MapBrowser.aspx (Build 37) which does the interpolation. In the drop-down box, select the correct chromosome and click the radio button for "physical positions (bp) only"
- Enter 30000000 and 50000000 on separate lines, then click "Get Map Positions".
- The output screen appears as:
No. Chr Query_Posn Sex_Ave_cM Female_cM Male_cM
1 1 30000000 56.0283 66.8713 47.2580
2 1 50000000 78.6422 99.8970 59.5195
- For sex-averaged value, subtract 56.0283 from 78.6422 (and round to 1 decimal point) = 22.6 cM
ICW vs. Triangulation (Source: Jim Bartlett, AdoptionDNA_Tools Message 2110)
- There is a big difference between ICW (which just means 3 people match anywhere - A matches B somewhere; A matches C somewhere; B matches C somewhere), and Triangulation (which means 3 people match each other on the same segment, and will necessarily have the same Common Ancestor).
Overlapping Segments, ICW, and TG (Source: Jim Bartlett, DNA-Newbie Message 40591)
3 related but separate terms (some paraphrasing by me):
- Overlapping segments – Shared segments that occupy the same space on a chromosome, which could either be on the maternal side or the paternal side.
- In Common With (ICW) – Means you and a match each match a 3rd person, somewhere – it could be the same segment is shared by all 3, or it could be on different shared segments (on the same or different chromosomes) between each pair of people out of the 3. Each of the 3 is ICW the others and not [necessarily – my addition] have the same Common Ancestor.
- Triangulated Group (TG) – Means all 3 people (or more) match each other (as in ICW) on the same segment. If 2 of your matches on an overlapping segment are ICW each other, the 3 of you form a TG 95% of the time.
- RootsWeb Genealogy-DNA Message 1402232409: An idea from Ann Turner for mapping the genome of an ancestral couple using segments confirmed by multiple lines of descent, with a link to her chromosome map for a line of her ancestors, with each of the pairs being compared are cousins of various degrees
A Trick for Adoptees on AncestryDNA
- You can link your DNA to trees you create from your close matches (have to do this one at a time). Select your closest match and build out a family tree for them (and all possible collateral lines) on Ancestry.com. [IMPORTANT: Ensure the tree is private and unsearchable. To do this, go to Tree Settings (which is to the right of your tree name), then Tree Settings, then Privacy Settings. Select the radio button for Private Tree, and also select the box for "Also prevent your tree from being found in searches".] Then attach your DNA results by going to your DNA Home Page, click on Settings, and link your DNA results to your close match on the new tree you have created. Then after a short while (a few hours) you may get some leaf hints.
How AncestryDNA Does Phasing
- Note that the thresholds for matching on 23andMe for X-chromosome DNA are different depending on the gender of the 2 individuals being compared (Source: ISOGG Wiki):
- X (male vs male): 200 SNPs, 1 cM
- X (male vs female): 600 SNPs, 6 cM
- X (female vs female): 1200 SNPs, 6 cM
- A response from Kathy Johnston (6 Jan 2015) on the ISOGG Facebook page (original message 5 Jan 2015) indicates she has observed an area of pileups on the X-chromosome from 70 to 80 million (not published)
|Defined by:||SNP||Set of STR values|
|Typical designation:||I2a1b2, R1b1a2a1a1b4*||13, 23, 14, 11, 14, 12, etc.|
|or by terminal SNP:||I-L161.1, R-L21||–|
|Defines ancestry from:||>100,000 years ago||~3,000 years ago|
|to (current use):||~1,000-3,000 years ago**||Present day|
|Subdivision known as:||Subclade||–|
|Mutations result in:||New SNPs forming new subclades||Changes to STR values at DYS|
|Mutations reversible:||Almost never||Yes|
* May vary from one organization to another (e.g., FTDNA vs. ISOGG)
** In the next few years the rigid definition given by SNPs will bring us into the second millenium
Nomenclature: DYS = DNA Y-chromosome Segment; SNP = Single Nucleotide Polymorphism; STR = Short Tandem Repeat
Attribution: Presentation by Nigel McCarthy, DNA Profiling of McCarthy Septs and Agnomen, at Genetic Genealogy 2013 (accessed 9 Nov 2013 via YouTube video http://www.youtube.com/watch?v=3sXEilMmnV4, but this video is now not available)
- How do I tell how closely I am related to a Y-Chromosome DNA (Y-DNA) match? What is FTDNATiP? – check out the link (for FTDNA Customers)
- Conversion Table for Y Chromosome Haplogroups (Wikipedia) – with all the changes in nomenclature of the Y Tree, this is a helpful link
This useful tool converts the SNP name to the longhand haplogroup name (for all the names from 2006). [Thanks to Debbie Kennett for posting this link on the DNA-Newbie List.]
Fixing and Phasing
"Fixing" Data for GEDmatch: a 23andMe Community thread that covers fixing no-call data using Felix Chandrakumar's DNA Error Fix tool and the file formats that need to be used. [Note to self to test out.]
Crossovers and Visual Phasing
- Split Segments Part 2 (Where did the Crossover Occur?) (accessible to those on the Yahoo! DNA-Newbie Group, Message 46616, 9 Jan 2015)
- Where Did the Crossover Occur? Use of siblings to match deceased grandparents through GEDmatch (accessible to those on the Yahoo! DNA-Newbie Group, Message 46863, 15 Jan 2015)
- Segment Matches with Grandparents Using Crossover Lines (accessible to those on FTDNA Forums, Message 36812, 23 Jan 2015) – has links to graphics (Visual Phasing of a Single Chromosome – the Use of Crossover Lines) and instructions (Instructions for Solving the Crossover Puzzle – Phasing and Matching to Grandparents)
- Crossover Mapping Instructions for 3 or More Siblings from Kathy Johnston (accessible to those on the Yahoo! DNA-Newbie Group, Message 47786, 8 Mar 2015), but note there can be "hidden coincidental crossovers", which are readily apparent when there is a switch from a FIR to no matching, but much more difficult to spot when hidden in a non-matching region. I love this technique, especially now that almost 90 of my relatives have undergone DNA testing. Although relatively easy for some chromosomes, it can be extremely challenging and frustrating on others; and if you guess, be prepared to be totally wrong!
- See this method posted (with Kathy's permission) on Jason Lee's site (20 Jan 2016) The Use of Crossover Lines to Determine Segment Matches with Grandparents Among Siblings — Visual Phasing and Joe Hartley's blog posting using this technique atDNA Under the Hood: Segments and Crossovers (5 Jan 2016)
- Take a look at links at the ISOGG Wiki: Visual Phasing
- YouTube: Visual Phasing using Excel – a great video by Lars Martin. Note that there is no sound and the menus are in Swedish, but for anyone familiar with Excel, it is very easy to follow. I have been using this same method using Excel for a while, but I include additional "floating columns" to allow for multiple, almost coincidental crossovers, which aren't immediately apparent, as trying to add them later throws everything off.
Average number of crossovers:
- Rule of Thumb for Approximate Number of Crossovers Per Chromosome from Jim Bartlett (accessible to those on the Yahoo! DNA-Newbie Group, Message 47953, 16 Mar 2015): Divide the total cM in a chromosome by 100, e.g., for chromosome 6, which is 194 cM on average has ~2 crossovers per generation, or for the total of 3587 cM for all 22 chromosomes, would expect about 36 crossovers in total (for one of the pairs of chromosomes)
- Escape from Crossover Interference Increases with Maternal Age, Campbell et al., 2014, Nature Communications 6: 6260 doi:10.1038 / ncomms7260 – Mean of 41.6 autosomal recombination events per gamete in females (95% CI: 41.4-41.9) and 26.6 in males (95% CI: 26.5-26.7)
- Genetic Analysis of Variation in Human Meiotic Recombination, Chowdury et al., 2009, PLoS Genetics 5(9): e1000648. doi:10.1371/journal.pgen.1000648
- The Recombination Project: Analyzing Recombination Frequencies Using Crowdsourced Data (Blaine Bettinger, The Genetic Genealogist, 20 Feb 2017)
- Age-Dependent Recombination Rates in Human Pedigrees, Hussin et al., 2011, PLoS Genetics, DOI: 10.1371/journal.pgen.1002251
DNA Testing Companies
- Apparently 23andMe does have "low-spitter" kits for $25 more if a tester is not able to produce sufficient saliva (Source: CeCe Moore, AdoptionDNA Mailing List #57321). My father, who is old and frail had difficulty in producing sufficient saliva using the standard kit – he was only able to produce about half of the required volume, but 23andMe was able to extract sufficient DNA from this.
- Despite a very professional YouTube video (GPS DNA Tracking) and a paper in Nature Communications (Geographic Population Structure Analysis of Worldwide Human Populations Infers their Biogeographical Origins, Elhaik E et al., Nature Communications, 29 Apr 2014), if you are thinking of forking out money for Prosapia Genetics' GPS (Geographical Population Structure) analysis ("Our first tool, GPS, will tell you where your DNA was forged, and is accurate to home village with a time resolution of the past 1,000 years"), I suggest you read Driving in the Wrong Direction with a Dodgy DNA SatNav (Debbie Kennett, 3 May 2014), including all the comments.
- The McGuire Method – Simplified Visual DNA Comparisons (Guest post by Lauren McGuire on The Genetic Genealogist, 19 Mar 2017) – This presentation technique for visualizing the relationships between testers and the amounts of shared DNA is fantastic. I really like Skip Duett's charts using this technique (Permalink from Genetic Genealogy Tips and Techniques closed Facebook Group). Lucidchart (free version that allows 3 active presentations and limited to 60 data elements + subscription-based for more complex presentations), a web-based flowchart maker, has been recommended as one of the diagramming tools for the McGuire Method.
Coefficient of Relationship
- Calculation of the Coefficient of Relationship by F. M. Lancaster (Oct 2005, updated Feb 2016) – see also Index. [Note: the original URLs no longer work, so the links provided are from the Wayback Machine.]
- How to Calculate the Coefficient of Relatedness (T.R. Smith, updated 21 Dec 2016). She provides the following formula:
The coefficient of relatedness (CR) between you and your Nth cousin M times removed is given by the formula
CR = (1/2)^[2N + M + 1]
assuming there is no multiple relationship or half relationship in the tree. Also, there is an interesting
- Wikipedia: Coefficient of Relationship – this has an interesting Figure.
- ISOGG Wiki: Coefficient of Relationship
- "The probability of having DNA from all of your genealogical ancestors at a particular generation becomes vanishingly small very rapidly; there is a 99.6% chance that you will have DNA from all of your 16 great-great grandparents, only a 54% of sharing DNA with all 32 of your G-G-G grandparents, and a 0.01% chance for your 64 G-G-G-G grandparents. You only have to go back 5 generations for genealogical relatives to start dropping off your DNA tree."
- "The number of genetic ancestors starts off growing exponentially, but eventually flattens out to around 125 (at 10 generations, 120 of your 1024 genealogical ancestors are genetic ancestors)."
- "There is an interesting effect of the larger recombination rate in women; you are, on average, slightly more closely related to your maternal line (your maternal grandmother, your mother’s maternal grandmother, etc) than you are to your paternal line (your father’s paternal grandfather, etc). You are about 30% more likely to be genetically related to your maternal-line ancestor 10 generations ago than you are the corresponding paternal-line ancestor (14% vs 11%)." [FASCINATING!]
I'm not wanting to put anyone off undergoing testing, but I love this quote from Kelly Wheaton (see Here):
People turn to DNA testing for different reasons and unfortunately learn too soon that their expectations are generally greater than DNA's testing ability to deliver just what they wanted. So first of all, if you haven't tested yet I would suggest lowering your expectations a few notches.
- Message 1439570335 – I agree with this list from Shannon Christmas
- Message 1439529915 – Tim Janzen's top 3
Some useful comments about 1:1 comparisons on GEDmatch and SNP thresholds: Rootsweb DNA-Genealogy message 1439670852
- Recommendations from Ann Turner for splitting files >1,048,576 rows (Excel's maximum): DNA-Newbie Message 50767
My tips for finding Ancestry and 23andMe profiles from a username or e-mail on GEDmatch
I responded to a message on the DNA-Newbie message board (43639):
You asked “I see that several of my stronger matches on Gedmatch were tested at Ancestry. Unfortunately, I have not found a good way to figure out what their profile is at Ancestry. Do you have any tips for matching profiles on Ancestry and Gedmatch?”
I often use Jeff Snavely’s extension for Google Chrome (AncestryDNA Helper, available Here), which has a search function for Users (the box above the buttons for the Hints, New, and Starred filters). Not infrequently, the name someone uses on GEDmatch is the same as their Ancestry name, 23andMe, or FTDNA name – or the first part of their e-mail is the same as their Ancestry/23andMe name. Searching on a first name (even a common one) using AncestryDNA Helper can be productive, with a very manageable list to go through. I also often run a check on an AncestryDNA match’s name in the User search box, to find other people matching me that the person administering the kit is managing – occasionally they have a Public tree (whereas the match I’m interested in either doesn’t have a tree or it is private). I’ve also had success in finding someone’s GEDmatch # by searching on my one-to-many match list for the first part of their e-mail on FTDNA, or their username on AncestryDNA or 23andMe – meaning I can still work on their segment data even when they won’t respond to messages or “share genomes” invitations.
The other thing I often do on GEDmatch is run a one-to-many for my match of interest. That then shows if they have any other close relatives on GEDmatch – and some of these people may show their e-mail addresses, even though the match I’m interested in doesn’t. For their close relatives, I also run GEDmatch comparisons to work out how they are probably related. And sometimes you can find an Ancestry tree (by searching using the AncestryDNA Helper extension as indicated above) for one of the closer relatives, even though your search was unsuccessful for the match of interest.
- AncestryDNA's Chromosome Numbering Conventions – an explanation of what AncestryDNA means when they report results for chromosomes 23, 24, and 25 (don't worry, we still only have 22 pairs plus the sex chromosomes!)
- Identical by State or Identical by Descent? by Ann Turner, Journal of Genetic Genealogy 2011 – this was a very illustrative article for me for IBD (identical by descent), identical by state (IBS), and compound segments, with examples of runs of alleles to illustrate each. An example was also given showing how a segment that matches both the child and the parent based on genotypes is not conclusive evidence for IBD, and that one child had 17/70 matches not found in either parent (with IBS lengths of 7.7-12.9 cM – and those with a lot of us with those "4th cousin" matches know that this length of segment is very common).
- GEDmatch.com's Phasing Tool (The Genetic Genealogist, Blaine Bettinger)
- Methods for Phasing (ISOGG Wiki) – Phasing is the process of trying to determine which DNA segments (autosomal or X-chromosome) were inherited from the father and which from the mother
- Phasing the Chromosomes of a Family Group When One Parent is Missing, T. Whit Athey, Journal of Genetic Genealogy 2011;6(1) (free PDF download)
- Chromosome Mapping for Genetic Genealogy (Tim Janzen)
- Genome Research article (free) by Pereira et al. (2005), High-resolution mtDNA evidence for the late-glacial resettlement of Europe from an Iberian refugium, also available as PDF: Genome Res. 2005. 15: 19-24
- BMC Evolutionary Biology article (free) by Brandstätter et al. (2008), Timing and deciphering mitochondrial DNA macro-haplogroup R0 variability in Central Europe and Middle East: BMC Evolutionary Biology 2008, 8:191
- Links for Creation of Phylogenetic Charts (McGee, Phylip, and MEGA Software) – from HAM Surname DNA Project
- Links to all the ISOGG Newsletters, which were published from 2008-2012
- A Beginner's Adventures in Genetic Genealogy (a work in progress) and Measuring the Length, the Rarity and the Relevance of Shared Autosomal DNA (Paddy Waldron) – I like Paddy's writing style and mathematical approach and the fact that he indicates up front, "I assume a knowledge of the basics and concentrate on the major flaws in autosomal DNA-matching as currently practiced. I will continue to play devil's advocate and to pick holes in the methodologies being used by the main DNA companies, by the third-party websites and by many genetic genealogists".
- The Trouble with Triangulation: Preliminary Notes (Ann Turner, 4 Apr 2015)
- How Much of Your Family Tree Do You Know? And Why Does That Matter? (Blaine Bettinger, The Genetic Genealogist, 11 Aug 2015)
- A Genealogical Look at Shared Ancestry on the X Chromosome (3 Apr 2016, Buffalo V, Mount SM, Coop G) – free download available at http://dx.doi.org/10.1101/046912
Page updated 3 April 2017