Biology Reference
In-Depth Information
Many computer programs now exist to construct linkage and QTL
maps. However, the i eld was brand new when we were mapping pollen
hoarding. We needed a program called MapMaker QTL that was devel-
oped by Eric Lander and his associates. Lander was leading the part of
the Human Genome Project that was working to develop tools for con-
structing genetic maps and mapping QTLs. He visited UC Davis about
the time we needed to begin building the map, installed his program
on my computer, and showed us how to use it. Although the mathemat-
ical algorithms for mapping are complex, the general principle is easy.
h ink about doing an independent statistical test at each marker along
one of the linkage groups shown in Figure 6.1. Start with marker J20.53
at the top of linkage group I and take all colonies that inherited the
high-strain allele for that marker on one list and all those that inherited
the low-strain marker allele on a dif erent list. h en ask the question
“Do the colonies represented on the two lists dif er from each other for
the amount of pollen stored in the comb?” If you do this successively
for every marker, you are scanning the genome for regions that seem to
be af ecting the amount of stored pollen. However, you are doing hun-
dreds of individual comparisons of markers that are not completely
independent of one another because they are genetically linked. QTL
mapping programs control for the large number of comparisons and
ge ne tic linkage.
Greg Hunt ran the program and found two major QTLs (Figure 6.3).
We had succeeded where I had been certain we would fail. Perhaps the
genetic basis of observed variation in naturally occurring social behav-
ior wasn't that complex at er all. We named the QTLs
pln1
and
pln2.
Together they explained 59 percent of the total variance in the pollen-
hoarding phenotype. Contrary to what we expected, the low-strain al-
lele for
pln1
resulted in more stored pollen than did the high-strain al-
lele. h is is a phenomenon that has persisted throughout our mapping,
across multiple individual phenotypes, and is dii cult to explain, but it
occurs commonly in plant mapping studies where two selected popula-
tions are crossed.
