MLINK: impossible result?

Frank Visser fvisser at
Fri Aug 18 07:58:16 EST 1995


I received a query from a user regarding some results he got when doing a 
two-point linkage analysis. Here follows the query:


              I have a problem with a simple two point linkage analysis. 
I have done many such analyses in the past and experience has taught me 
to expect that, with any fully penetrant dominant phenotype, a marker 
that detects a crossover will always give negative infinity at a theta 
of zero. However I now have a family with about 40 meioses in 60 or so 
individuals in which this is not happening. Is this possible? The single 
crossover is between branches of the family separated by seven unsampled 
meioses (the people involved are dead) but it is quite unambiguously 
there. The disease segregates with one allele in all the rest of the 
family and with another in a small branch. This is only clearly visable 
in two individuals but nevertheless it is there.The lod score in 
question rises to a peak of over 5 at about three centimorgans then 
falls to 3.4 at zero. I've checked it, I've run it on both a PC and the 
HGMP facility, I've played with allele frequencies of both the linked 
marker and the disease, I've run both two point (MLINK) and multipoint 
(LINKMAP), I've even typed the whole thing in again from scratch, and 
still it is there. The lod score goes up and down, but is always still 
posative at a theta of zero. Have I made a mistake, or is this a real 
result? Any help or advice would be gratefully recieved.

So, is it possible to get such a result for a fully penetrant dominant 
phenotype? Or is more detailed knowledge of the pedigree 
necessary? Is 
something wrong with the model? I tried to figure it out for myself 
but am not sure. Intuitively I would say that something is wrong, but I 
do not have any idea where to start to look. Any information, discussion, 
pointers to literature, etc. gladly welcomed. All information will be passed 
on to the user. 

F. R. Visser 
HGMP-Resource Centre
Hinxton, UK
Phone : +44 (0)1223 494526
e-mail: fvisser at


More information about the Gen-link mailing list