Hi Again,
I'm continuing with more material from Ralph Taylor on the meaning of matches. What does it mean if you match 67 out of 67, commonly written 67/67, with another person? There is about a 50% likelihood that you share the same great-grandfather! Thus, you are second cousins. You could be slightly closer related or slightly more distantly related, but you are related recently as genealogists measure relationship!
As Ralph Taylor says, "67-marker matches, if you’re lucky enough to find one, are very precise; you may need to account for the birth dates of the matching individuals. The main problem is that most people have done a lesser test, so you can only compare against the markers they had tested." Thus, refer to earlier blogs which discuss 37 marker matches and fewer matches.
Should you match in the range of 65-67/67: You share a common ancestor and he almost certainly lived within a genealogic time frame.
o A perfect 67/67: The 90% probability level is a TMRCA of <= 11 generations, >= 1735 AD; 70% is <= 6 gen., >= 1860 AD; & it’s 50% probable that your common ancestor is your great-great-grandfather.
o 66/67: 90% is <= 18 gen., >= 1560 AD; 70% is <= 18 gen., >= 1730 AD; 50% is <= 7 gen., 1835 AD.
o 65/67: 90% is <= 24 gen., >= 1410 AD; 70% is <= 16 gen., >= 1610; 50% is <= 12 gen., >= 1710 AD.
Should you match in the range of 62-64/67: You share a common ancestor, who may or may not have lived in a genealogic time frame.
o 64/67: 90% is <= 31 gen., >= 1235; 70% is <= 22 gen., >= 1460; 50% is <= 16 gen., 1610.
o 63/67: 90% is <= 37 gen., >= 1085 AD; 70% is <= 27 gen., >= 1330 AD; 50% is <= 21 gen., >= 1480 AD.
o 62/67: 90% is <= 43 gen., >= 930 AD; 70% is <= 32 gen., >= 1200 AD; 50% is <= 26 gen., >= 1360 AD.
We've have spent a number of blog entries discussing the statistical meaning of matches. You do not need to completely understand statistics or probability, though. What you do need to do is to investigate the family trees of anyone who closely matches you. For example, with a "perfect" match of 67/67, you and the other person should be able to find your common ancestor fairly easily. One hopes that each of you has some information about your mutual family tree so that you each gain significant new information. After all, that's why you chose to participate in this project.
The good news is that eventually sufficient y-DNA records and pedigrees will exist to satisfy every participant's desire for knowledge of their roots!
Best wishes as we move towards that goal.
Lalia
Showing posts with label statistics. Show all posts
Showing posts with label statistics. Show all posts
Friday, September 4, 2009
Thursday, August 20, 2009
37-Marker Matches and What They Mean
Hi,
What happens when you test 37 markers? Does a match of 36 out of 37 markers, commonly abbreviated as 36/37, mean more than 23/24?
Here is some more information from Ralph Taylor on what matches mean when we have 37 loci or markers to consider.
37-marker matches get even more precise. For example, if you share 35, 36 or 37 markers with another Taylor (or someone with your surname, if you are not looking at Taylor project results!) then you share a common ancestor, probably within a genealogic time frame.
o A perfect 37/37: The chances are 90% that your TMRCA is <= 19 generations, >= 1530 AD; 70% prob. is <= 10 gen, >= 1760 AD; 50% is <= 6 gen., >= 1860 AD.
o 36/37: 90% is <= 32 gen., >= 1210 AD; 70% is <= 20 gen., >= 1500 AD; 50% is <= 14 gen., >= 1660 AD.
o 35/37: 90% is <= 45 gen., >= 885 AD; 70% is <= 30 gen., >= 1260 AD; 50% is <= 22 gen., >= 1460 AD.
However when you have 37 markers to compare, and you don't match 35 or more, such as 33-34/37: You share a common ancestor, but probably not within a genealogic time frame.
§ 34/37: 90% is <= 57 gen., >= 580 AD; 70% is <= 40 gen., >= 1000 AD; 50% is <= 31 gen., >= 1235 AD.
§ 33/37: 90% is <= 69 gen., >= 285 AD; 70% is <= 50 gen., >= 760 AD; 50% is <= 40 gen., >= 1010 AD.
When we get to 32 matches out of 37, or worse: You may share a common ancestor, but the probability that he lived in within a genealogic time frame is small. The 50% probability level for a 32/37 match is <= 51 generations, >= 735 AD.
For some of you, these discussions of probability and statistics are dull. However, they are necessary to understand the meaning of genetic genealogy. We will have future posts with specific examples. For now, consider that your surname is Taylor. You have a family tree with some degree of confidence. Your fourth cousin, another male Taylor descended from the same line also tests. Now, if you match 37/37 or 36/37, this suggests that the family tree is correct for both of you. If, however, your match is 32/37 or less, the family tree is not correct for one or both of you. Because two males from the same male line, only separated by 6 generations would not differ to this extent.
May all your genealogical puzzles be solved!
Lalia
What happens when you test 37 markers? Does a match of 36 out of 37 markers, commonly abbreviated as 36/37, mean more than 23/24?
Here is some more information from Ralph Taylor on what matches mean when we have 37 loci or markers to consider.
37-marker matches get even more precise. For example, if you share 35, 36 or 37 markers with another Taylor (or someone with your surname, if you are not looking at Taylor project results!) then you share a common ancestor, probably within a genealogic time frame.
o A perfect 37/37: The chances are 90% that your TMRCA is <= 19 generations, >= 1530 AD; 70% prob. is <= 10 gen, >= 1760 AD; 50% is <= 6 gen., >= 1860 AD.
o 36/37: 90% is <= 32 gen., >= 1210 AD; 70% is <= 20 gen., >= 1500 AD; 50% is <= 14 gen., >= 1660 AD.
o 35/37: 90% is <= 45 gen., >= 885 AD; 70% is <= 30 gen., >= 1260 AD; 50% is <= 22 gen., >= 1460 AD.
However when you have 37 markers to compare, and you don't match 35 or more, such as 33-34/37: You share a common ancestor, but probably not within a genealogic time frame.
§ 34/37: 90% is <= 57 gen., >= 580 AD; 70% is <= 40 gen., >= 1000 AD; 50% is <= 31 gen., >= 1235 AD.
§ 33/37: 90% is <= 69 gen., >= 285 AD; 70% is <= 50 gen., >= 760 AD; 50% is <= 40 gen., >= 1010 AD.
When we get to 32 matches out of 37, or worse: You may share a common ancestor, but the probability that he lived in within a genealogic time frame is small. The 50% probability level for a 32/37 match is <= 51 generations, >= 735 AD.
For some of you, these discussions of probability and statistics are dull. However, they are necessary to understand the meaning of genetic genealogy. We will have future posts with specific examples. For now, consider that your surname is Taylor. You have a family tree with some degree of confidence. Your fourth cousin, another male Taylor descended from the same line also tests. Now, if you match 37/37 or 36/37, this suggests that the family tree is correct for both of you. If, however, your match is 32/37 or less, the family tree is not correct for one or both of you. Because two males from the same male line, only separated by 6 generations would not differ to this extent.
May all your genealogical puzzles be solved!
Lalia
Friday, August 14, 2009
25-Marker Matches
What happens if/when we have 25-marker matches?
Well, the good news is that 25-marker matches allow finer interpretations:
· 24-25/25: You share a common ancestor, probably within a genealogic time frame. (The numbers of generations & dates below are approximate.)
¨ A perfect 25/25: The chances are 90% that your most recent common ancestor is within the past 28 generations, roughly since 1300 AD. It’s 70% that the time to the most recent common ancestor (TMRCA) is within 15 generations >= 1630 AD and 50% the TMRCA is <= 9 gen., >= 1780 AD. Thus, as we will find out when we look at greater numbers of markers, discovering a "perfect" 25/25 match only means that you share an ancestor who lived about 1300 (90% probability) to 1780 (50% probability).
A 24/25 match means even less. 24/25: The 90% probable TMRCA is <=48 gen, >= 800 AD; 70% is <= 30 gen., >= 1260 AD; 50% is <= 21 gen., >= 1480 AD.
Matching 23/25 may not illuminate your genealogical studies at all, as your common ancestor may have preceeded records with surnames. 23/25 or worse: You share a common ancestor, but he may not have lived within a genealogic time frame
Looking at this in terms of generations, instead of years, the match of 23/25: 90% TMRCA is <= 67 gen., >= 330 AD; 70% is <= 46 gen., >= 860; AD; 50% is <= 34 gen., >= 1160 AD.
While a 22/25 match means even less. 22/25: 90% TMRCA is <= 86 gen., >= 140 BC; 70% is <= 61 gen., >= 480 AD; 50% is <= 47 gen., >= 830 AD.
Input from Ralph Taylor and Lalia Wilson
Well, the good news is that 25-marker matches allow finer interpretations:
· 24-25/25: You share a common ancestor, probably within a genealogic time frame. (The numbers of generations & dates below are approximate.)
¨ A perfect 25/25: The chances are 90% that your most recent common ancestor is within the past 28 generations, roughly since 1300 AD. It’s 70% that the time to the most recent common ancestor (TMRCA) is within 15 generations >= 1630 AD and 50% the TMRCA is <= 9 gen., >= 1780 AD. Thus, as we will find out when we look at greater numbers of markers, discovering a "perfect" 25/25 match only means that you share an ancestor who lived about 1300 (90% probability) to 1780 (50% probability).
A 24/25 match means even less. 24/25: The 90% probable TMRCA is <=48 gen, >= 800 AD; 70% is <= 30 gen., >= 1260 AD; 50% is <= 21 gen., >= 1480 AD.
Matching 23/25 may not illuminate your genealogical studies at all, as your common ancestor may have preceeded records with surnames. 23/25 or worse: You share a common ancestor, but he may not have lived within a genealogic time frame
Looking at this in terms of generations, instead of years, the match of 23/25: 90% TMRCA is <= 67 gen., >= 330 AD; 70% is <= 46 gen., >= 860; AD; 50% is <= 34 gen., >= 1160 AD.
While a 22/25 match means even less. 22/25: 90% TMRCA is <= 86 gen., >= 140 BC; 70% is <= 61 gen., >= 480 AD; 50% is <= 47 gen., >= 830 AD.
Input from Ralph Taylor and Lalia Wilson
Saturday, August 8, 2009
What Does a 12 Marker Match Mean?
Here are some thoughts from Ralph Taylor on 12-marker Matches:
About 12-marker matches, there isn’t a lot to say. There are only three possibilities:
A perfect 12/12 means you share a common ancestor. But that ancestor may be many centuries in the past. The 90% probability level includes the past 60 generations (dating to as early as 510 AD); the 70% level includes the past 31 generations (as early as 1235 AD; and the 50% level (as likely to be low as high) includes the past 18 generations (as early as 1560 AD).
11/12 is inconclusive. Comparison of more markers is needed to draw any conclusions. If you are the one with only 12 markers tested, you should upgrade to more markers.
10/12 or worse excludes sharing a common ancestor within several thousands of years.
That’s cold, but true. 12 markers are considered sufficient to exclude sharing of a common ancestor, but not to include it.
Now, Lalia's interpretation... This means quite simply that if you match someone with the same surname 12 for 12, it suggests that you share that surname. However to trace your ancestry, you are going to need to test more markers and match more markers! (About which we will have more information in a future blog.)
If you do not share the same surname, or a variant of the surname, then a 12/12 match means that you likely come from the same general gene pool, such as the British Isles, and means nothing more.
We will have some forthcoming blog posts about higher levels of matches.
Good luck with your own genetic genealogical projects!
Lalia
About 12-marker matches, there isn’t a lot to say. There are only three possibilities:
A perfect 12/12 means you share a common ancestor. But that ancestor may be many centuries in the past. The 90% probability level includes the past 60 generations (dating to as early as 510 AD); the 70% level includes the past 31 generations (as early as 1235 AD; and the 50% level (as likely to be low as high) includes the past 18 generations (as early as 1560 AD).
11/12 is inconclusive. Comparison of more markers is needed to draw any conclusions. If you are the one with only 12 markers tested, you should upgrade to more markers.
10/12 or worse excludes sharing a common ancestor within several thousands of years.
That’s cold, but true. 12 markers are considered sufficient to exclude sharing of a common ancestor, but not to include it.
Now, Lalia's interpretation... This means quite simply that if you match someone with the same surname 12 for 12, it suggests that you share that surname. However to trace your ancestry, you are going to need to test more markers and match more markers! (About which we will have more information in a future blog.)
If you do not share the same surname, or a variant of the surname, then a 12/12 match means that you likely come from the same general gene pool, such as the British Isles, and means nothing more.
We will have some forthcoming blog posts about higher levels of matches.
Good luck with your own genetic genealogical projects!
Lalia
Terminology Used in Genetic Genealogy
Hi,
Here are some terms that Ralph Taylor provides to help you understand your reading of articles and blog posts about genetic genealogy.
MRCA: The most recent common ancestor shared by two or more persons with a DNA match. We are concerned only with the most recent, because preceding ancestors will also be shared.
TMRCA: The probable time to the MRCA, expressed in generations.
Probability level: The cumulative probability, expressed as a percentage, that a MRCA lived within the specified number of generations. Here, we give three probability levels, 90%, 70%, & 50%. 90% means the chances are 9/10 the TMRCA is <= the given number; 50% probability is as likely to give a low TMRCA as a high one.
Genealogic time frame: A time for which it is feasible to specifically identify an ancestor, generally from about 1350 AD to the present.
Generations, length of: This is an estimated average of 25 years, developed from 18th and 19th century America statistics. We use the estimate to convert generations to common dates. Other societies may have had a different generational length -- either longer or shorter.
<=: Means less than or equal to, the maximum generations number for that probability level
>=: As applied to dates, means no earlier than the given date.
Here are some terms that Ralph Taylor provides to help you understand your reading of articles and blog posts about genetic genealogy.
MRCA: The most recent common ancestor shared by two or more persons with a DNA match. We are concerned only with the most recent, because preceding ancestors will also be shared.
TMRCA: The probable time to the MRCA, expressed in generations.
Probability level: The cumulative probability, expressed as a percentage, that a MRCA lived within the specified number of generations. Here, we give three probability levels, 90%, 70%, & 50%. 90% means the chances are 9/10 the TMRCA is <= the given number; 50% probability is as likely to give a low TMRCA as a high one.
Genealogic time frame: A time for which it is feasible to specifically identify an ancestor, generally from about 1350 AD to the present.
Generations, length of: This is an estimated average of 25 years, developed from 18th and 19th century America statistics. We use the estimate to convert generations to common dates. Other societies may have had a different generational length -- either longer or shorter.
<=: Means less than or equal to, the maximum generations number for that probability level
>=: As applied to dates, means no earlier than the given date.
Saturday, July 25, 2009
Taylors in the USA
If you are interested enough in Taylors to be reading this blog, you’ve probably already searched for Taylors in public records. Not an easy task, is it? Then, if you combine the Taylor surname with a common first name, say John or James, you are in real trouble. You can find many of them even in the same counties you know your ancestor of the same name lived…
But here’s proof of how many Taylors are out there. This information was obtained from Ancestory.com’s census records and then corrected, because a number of censuses underreported in some years. Thanks to Ralph Taylor and Josh Taylor, project administrators who did the hard work.
Consider these numbers:
US Federal Census Data for Taylor, 1790-1930
From search on "taylor", exact match, exact spelling at Ancestry.com US federal census indices, with corrections for underreporting in some jurisdictions.
1790 9,710
1800 13,300
1810 17,940
1820 23,600
1830 33,650
1840 58,560
1850 67,430
1860 85,620
1870 133,270
1880 171,360
1890 211,160
1900 248,515
1910 283,066
1920 313,925
1930 341,174
So, how do you find the right Taylor ancestors? Glad you asked that. Y-chromosome DNA is one way to go about locating the right Taylor family. Go to www.familytreedna.com and look for the Taylor surname project. We are now at 277 members and just beginning to fully elucidate the Taylor gene pool.
And, yes, there are other ways to find your correct Taylor ancestors. Go for the paper trail with public and private documents. And, realistically, you need to do both. The combination of biological proof and a paper trail will make definite proof that you are investigating the right ancestors.
Good luck with your genealogy research! Remember to get back to the Taylor surname project when you have updated your Taylor family tree.
Lalia
But here’s proof of how many Taylors are out there. This information was obtained from Ancestory.com’s census records and then corrected, because a number of censuses underreported in some years. Thanks to Ralph Taylor and Josh Taylor, project administrators who did the hard work.
Consider these numbers:
US Federal Census Data for Taylor, 1790-1930
From search on "taylor", exact match, exact spelling at Ancestry.com US federal census indices, with corrections for underreporting in some jurisdictions.
1790 9,710
1800 13,300
1810 17,940
1820 23,600
1830 33,650
1840 58,560
1850 67,430
1860 85,620
1870 133,270
1880 171,360
1890 211,160
1900 248,515
1910 283,066
1920 313,925
1930 341,174
So, how do you find the right Taylor ancestors? Glad you asked that. Y-chromosome DNA is one way to go about locating the right Taylor family. Go to www.familytreedna.com and look for the Taylor surname project. We are now at 277 members and just beginning to fully elucidate the Taylor gene pool.
And, yes, there are other ways to find your correct Taylor ancestors. Go for the paper trail with public and private documents. And, realistically, you need to do both. The combination of biological proof and a paper trail will make definite proof that you are investigating the right ancestors.
Good luck with your genealogy research! Remember to get back to the Taylor surname project when you have updated your Taylor family tree.
Lalia
Subscribe to:
Posts (Atom)
