Clusters represent shared segments - jonathanbrecher/sharedclustering GitHub Wiki
This is an interesting example of a real life analysis, originally posted to the Shared Clustering User Group by Brian Schuck. This is a comples example of interpreting clusters with red overlap:
A pattern I've noticed when looking at the clusters of the various kits I analyze is something I'll refer to as "The Cross".
One can observe a heavy red vertical line and a less red horizontal line. My observation is that in this case, the matches in the upper left quadrant match on the left hand side of the segment in question (in this case nearly all of chromosome 19). Likewise the matches in the bottom right quadrant match on the right side of the chromosome. The segment lengths are long enough (40+ cM) so that it's possible to have 20 cM matches on either side of the chromosome, that don't show up as 20 cM matches to each other. Many of these matches have been uploaded to GEDmatch and other services so that a chromosome browser view of how this cluster looks was possible to create. The colors on the first picture correspond to the same colors on the second picture. It's remarkable how the clustering technique put these matches in an order that is nearly identical to how they lay out in the chromosome.
As an aside, these matches are actually quite distant and the majority match only on this one chromosome. The common ancestors were born prior to 1730 in Germany.
