This is the current “working tree” status of the Q-L11515 branch that has spread over much of northern Scandinavia. (Right klick and choose “view image” or “open image in new tab” to get it Zoomable). Green squares mean that we have at least two kits leading to the same forefather. The orange ones are of course important to get tested to confirm the correctness. The Q-L11515 branch i part of the big tree of paternal lines.
I have made some experiments with predicting major haplogroup branch (U5a1, U5a2, U5b1, U5b2, U5b3) using machine learning techniques. I have tested multinominal logistic regression (“multilogit”), random forest, discriminant analysis and K Nearest Neighbor. So far I have had best luck in classification accuracy with multinominal logistic regression so I will present those results. Only the HVR1 region is used as information and the models are estimated using those kits that have a definite haplogroup.
This is the geographical locations of the YFull kits of the interesting branch Q-L804 with closest relatives among the American indians. Some how the Q-L804 branch seems to have made it to northern Europe at some point, likely through Siberia.
Non YFull kits gets a highly transparent dot in the predicted branch colour that can show the dencity of Q-L804.
The Genographic Project (Geno) kits with Q-L804 Geno stoped collecting samples in August 2019 but have captured some Q-L804.
This is the current status of the family tree of Q-L527 based on the data at YFull (Updated 2020-01-01). Two interesting kits from Belarus and Ukraine on old branches. They might give an indication of a possible migration route to Scandinavia. The main expansion in number of branches of Q-L527 seems to be before the viking age. I have estimated branch length in years with the Ape package in R using maximum likelihood and the “strikt model”.
Hans Ersson, 1655, illustrates the impact an individual can have on the genetic make up of a region. 25% of those born 1890-1930 in Sorsele had a strait paternal line to him according to the database Kråken. Quite extrem!
To see if there were more extrem cases I made a histogram of the number of “founding fathers” and their number of descendents in the database with a strait paternal line to them.
The maps illustrates the impact of the “Örträsk Finns” on the interior of Västerbotten by three maps showing how their descendants spread. It is based on the database Kråken which is not compleatly finished for all of Västerbotten but it can anyway give a rough picture of how they spread over the land. For illustration three of the founding fathers are used namely:
Johan Philipsson Hilduinen ca 1620-1697 Mårten Hindersson ca 1625-1697 Erik ca 1640 only known from his two sons Erik and Håkan The first map shows the proportion of the persons born 1890-1930 who descend from them.
Arctophilacius This is my paper link back to the big Sursill, Arctophilacius and Calamnius family. The link is probably wrong as there are conflicting DNA tests of the paternal lines from Petrus Arctophilacius at present.
See more about this family at:
Arctophile Origin of name???? 🐻 😄 A person who is very fond of and is usually a collector of teddy bears.
Arctophile means just “bear loving, bear lover,” but in modern English specifically a lover of teddy bears, not grizzlies.
I made a map of the YFull kits with province listed for the Scandinavian countries and Finland with the basic idear that one kit should be one point. Mainly as an experimental alternative to other maps for showing the geographical distributions of the haplogroups. To avoid over plotting each point searches for a free space to be plotted and can also adjust size to squese in. This causes the ball shapes in especially Finland as most points are located to province landmass centroides.
This is the current status of the family tree of Q-L804 based on the data at YFull. Branch length in years I have estimated with the Ape package in R using maximum likelihood and the “strikt model”. The rightmost codes refer to actuall test persons (YFxxxxx). YF02661 is a decendant of Erik Ersson in Knaften. YFulls version of the tree can be found at: https://www.yfull.com/tree/Q-L804/.
The idea of maximum likelihood here is to choose the branch lengh that gives the most likely “pedigree” given the observed number of SNPs at every branch.