Recombination occurs at different rates in fathers and mothers. This model produces highly accurate estimations of shared DNA.
These model results show the ranges for shared DNA between relatives and ancestors. It’s trained to reproduce three known standard-deviations from peer-reviewed science journals. It is currently being updated to reflect an average of 54 breaks per genome.*
Data resulting from the new changes will be shown below, then a clear separation will be made, and then data resulting from the old method will be shown below that.
Following these updates, I don’t foresee making very big changes to this model in the future—perhaps only some tweaks to the input number of recombinations for mothers and fathers if more accurate values become available for the standard deviations used to train the model.
I’ve already posted a comparison of this model to two previous, simpler models; but I want to provide an area where people can see the results of the most accurate model all in one place. I hope that this can assist people who are trying to identify how closely they might be related to a DNA match or figure out how usual or unusual their percentage of shared DNA is with a known relative.
Figure 1. Same as the cover photo. Results for shared DNA between siblings. Sib. = sibling. HIR = ‘half-identical regions,’ where one of the two chromosome homologues matches. FIR = ‘fully-identical regions,’ where both copies of a chromosome match. HIR + FIR = all of the points on chromosomes where two people match once plus all of the points where they match on both copies. A 3/4 sibling to you would be the result of one of your parents having a child with a sibling of your other parent. Results are from a two homologue autosomal Poisson model. Each row is the result of 500k simulation trials. Each recombination event has 20k gene segments available.
Figure 2. Results for grandparents or descendants of grandparents. All parameters are the same as for Figure 1.
Figure 3. Results for descendants of grandparents, continued, for half-relationships. All parameters are the same as for Figures 1 and 2.
Figure 4. Results for descendants of great-grandparents.Two homologue autosomal Poisson model. Each row is the result of 500k simulation trials. Each recombination event has 1700 gene segments available. This figure will be replaced by new simulation runs with 20k segments available.
Results below are for the old model parameters, where 1700 genes segments were available for recombination rather than 20k.
Figure 5. Results for 2nd great-grandparents. All parameters are the same as for Figure 4. This figure will be replaced by new simulation runs with 20k segments available.
Figure 6. Results for 3rd great-grandparents. All parameters are the same as for Figures 4 and 5. This figure will be replaced by new simulation runs with 20k segments available.
I hope you’ve found these results useful. More will be on the way.
*Changes are being made as of 14 September 2020. Previously, I had used 55 recombinations per genome, or 33 recombinations plus 22 chromosomes. Changing the number of breaks from 55 to 54 doesn’t change the results of the simulation significantly, but I was already planning to re-run the simulation: 20k segments are now being made available to split up during recombination, versus the 1700 that were previously used solely because of time and computational constraints. This was drastically assuaged by converting the code from Python to Java and upgrading hardware. I have recently simulated 20k, 50k, and 100k segments, only to find minuscule changes in the results between the three values.
Feel free to ask me about modeling & simulation, genetic genealogy, or genealogical research. To see my articles on Medium, click here. And check out my nifty calculator that’s based on the first of my three genetic models. It lets you find the amount of an ancestor’s DNA you have when combined with various relatives.