A Segment-ology TIDBIT…
Crossovers and segments are formed by random DNA biology. They are formed at conception in each of our ancestors and in ourselves. They are at fixed, permanent locations in each of us. They are not affected by family size, geography, wealth, status, intelligence, etc. For each of us, they are a fixed structure of our chromosomes – like a picture or jigsaw puzzle – which is different for each person.
22A Segment-ology: Crossover and Segment Formation TIDBIT by Jim Bartlett 20170101
segment-ology 
I sometimes have groups of segments with identical start and/or end positions so these are clearly crossovers. I also know segment ends are usually fuzzy. What I want to know (if possible) is if you can give me a ballpark estimate of what is a reasonable amount of variability to expect for a “real” crossover point. I often have groups of segments with pretty close end points on one or both ends.
LikeLike
Christine – You are correct that Triangulated Group segments are bounded (each end) by crossovers. And a crossover is an actual point between two base pairs, which we “see” as a close SNP – and the system has several factors which result in the beginning or end of shared segment as being a little fuzzy. So a consensus of start/end positions is one way to tell. But, in determining all the TG segments along one chromsome, we find where one TG ends another one starts. For me it’s generally more obvious where a segment starts, so I use that and just accept that the previous segments dribbles past that point some. However, two things to watch out for:
1. Closer Matches with larger segments may well span two (or more) other TG segments. Clearly our large grandparent segments (with 1Cs) will overlap several segments from more distant Ancestors.
2. Some smaller segments in a TG may be because they didn’t get the full segment we did (or alternatively, we didn’t get the full segment an Ancestor passed down); OR it may be that they actually came from a more distant Ancestor (past the Ancestor who formed the full TG segment)
I think the key is to determine all the segments on a chromsome and make a judgment call where one segment ends and another one starts.
Look at the colorful chromosome map at the top of all Segmentology pages for examples of adjacent segments. Jim
LikeLike
They are not formed at conception. They are already fixed in sperm and egg before conception. They are formed during gametogenesis.
LikeLike
Andrew, Point taken, thanks! In one sense many different combinations of crossovers and segments are formed before conception; but it is only at conception that you get a specific sperm and egg combination.
LikeLike
True, the specific combination is only determined by which egg and sperm unite. But if you want to understand the details, like inihibition whcih prevents crossovers being very close together, then you need to know about gametogenesis.
LikeLike
Andrew, I appreciate your input. I’m an old genealogist – my grandmother got me hooked in 1974. And I’m an Professional Engineer – now retired after 50 years of work. I just finished a free online (EdX) course (DNA: The Secret of Life) from MIT by Prof. Eric Lander – he was great, and I really had to concentrate to keep up. My wife has a PhD in microbiology, and I wanted to really get grounded. But for this blog, my focus is on genealogy – my desire is to keep it at a level that most genealogists (and I) can understand. For that reason, I try to stay away from, as much as I can, the details, and the language that goes with it. Most genealogists want to know how the DNA works to help them in their hobby. So the takeaway on this post was that the segments from our ancestors are fixed. We use the DNA to get the locations roughly right and then use genealogy to link the segments with the correct ancestors. And for the readers, we can point out that in one generation, the crossovers will not be close to each other. However, in a different generation a new crossover can occur very close to an existing one, giving the appearance of close crossovers. In other words, it’s possible to have some fairly small TGs in our chromosome map. Since we don’t usually work with very small shared segments, we probably won’t “see” these small TGs – they may appear as small gaps between larger TGs.
LikeLiked by 1 person
Do different companies show the same or different base addresses for a crossover? Does GEDmatch retain the same base address?
LikeLike
lkessler,
This is a two part question. (1) All the companies, generally, use the same base addresses for the same SNP. This makes the results from different companies compatible, even if they each have a different mix of SNPs. (2) No company knows the specific SNP of a crossover. They don’t even try. They look for long strings of matching SNPs between you and a Match, and report that. (This shared segment may or may not start or end on a crossover, even if the data were precise). Since there are only 4 values involved (A, C, G and T), and because you have these values on two chromosomes (and so does your Match), the matching algorithm can actually run way past the actual end of an ancestral segment and still find matching SNPs (at 23andMe and GEDmatch). At FTDNA, they use blocks of 100 SNPs – you will often see the length of a shared segment with the number of SNPs ending in 00. In this case the algorithm wants to see whole blocks match, and so the shared segment is often reported as shorter than it really is. This all results in fuzzy data! But it is all still in the ballpark. It still results in a hit on our broad TG “target”. It’s OK for genealogy. If you are looking for a specific gene near the end of a TG, you really need to be using phased data. Otherwise, for genealogy, we are “good to go.” Jim
LikeLike