DYS464

I just answered a question from a researcher about 464.  This is a confusing marker - so I've summarized the main aspects for anyone who needs a better insight.

DYS464 does not have actual sequence on the y-chromosome.  The values are reported in rising sequence - as a convention.  To make a comparison, you first line up the results to achieve the maximum number of matches and then place any that don’t match in the remaining position(s).  Then, you count the number of positions that don’t match to determine the Genetic Difference.  (Additionally - this marker is a special case where you don’t sum the absolute value of the differences)

Example: if you are comparing a 12,12,14,16 sequence to a 12,14,15,16 sequence, your first instinct is to see them in reported sequence and to conclude that here are two differences - one at 464b and one at 464c. 

12,12,14,16
12,14,15,16
 

However, the appropriate way to make this comparison is to rearrange one set to maximize the number of matches - so the comparison becomes:

12,12,14,16
12,15,14,16

Despite the fact that the mathmatical difference between 12 and 15 is 3, the Genetic Distance is considered to be 1 because of the volatility of this marker.  So, whether the second result was 12,14,15,16, 12,14,14,16 or 12,13,14,16 – the Genetic Distance is 1.

And - if your result is something like 12,12,16,16 while your matches are 12,13,14,16 - it's very possible that you are also seeing some "doubling" at other multi-markers, which is a special situation we abbreviate as "recLOH".  Rather than trying explain this complicated (and rare) phenomenon in this posting, contact me if you are seeing doubling at your multi-markers (385, 459, 464, YCAII, CDY, 395S1, 413 ...) and I'll help you with it.

DYS 464, Multi-Markers & recLOH: Great Subjects for Future Blogs

Hello Terry,

Thanks for clarifying the most useful way to compare the DYS464 markers.  I need to go back and reassess some of the 464 terrain in our project's result table.  It would be worthwhile to know if the other multi-markers should be compared by similar means.  The "recombinant loss of heterogeneity" issue is one that you mentioned to me some time ago, prompting me to read up on it.  It's an interesting subject that might be worth addressing in a blog entry all its own (or maybe a series of entries).

Thanks for all you do!

-Mark Bunch