Biology Reference
In-Depth Information
is the exponential decay parameter, and s
2
is a new relative
variance component that measures the relative proportion of variance
accounted for by the random spatial process. The term exp(
they live in,
l
) gener-
ates a valid spatially decaying positive correlation matrix that structures
the spatial variance component. As
lD
, this model becomes a stan-
dard shared household model in which individuals within a household
exhibit a correlation of 1 while across households the resulting correlation
is zero. This model can also be expanded to allow for host genet-
ic
l
/N
spatial interactions by adding an additional variance component that
is structured by the Hadamard product (i.e. the matrix operator that
involves element by element multiplication) of the two relevant covari-
ance kernels (i.e. covariance structuring matrices derived from either the
pedigree information or the spatial information).
Using this spatial model, we found that both a host genetic factor and
a random spatial component are required to best fit the data on Ascaris
worm burden.
64
We observed no evidence of host genetic
spatial
interaction. The observed estimate of the exponential decay parameter in
the general model was 3.92 which suggests that the expected spatially
varying correlation is halved by a distance of 0.177 km. In the general
model, host genetics accounts for approximately 27.2 (
9.6%) of the total
variation in Ascaris worm count while spatial factors account for an
additional 9.6% (
4.1%). Thus, host genetic factors still appear to be the
single largest determinant of phenotypic variation. The observed spatial
component may reflect the worm genetic component or some other
environmental spatial process but this still remains to be resolved.
Our interpretations for these more complex models are still equivocal.
Direct assay of between-worm genetic variation appears to show that
genetic variation within the parasite is not important for explaining host
variation in total worm burden. However, our much larger approximate
study using spatial variation strongly suggests that some type of spatial
decay process is important. Such decay may still be due to underlying
worm genetics or may simply be due to spatial variation in exposure that
would represent a case where environmental sharing also influences the
phenotypic correlations among individuals.
ASCARIS
INFECTION: DEEP SEQUENCING TO IDENTIFY
RARE FUNCTIONAL VARIANTS
FUTURE OF GENETIC RESEARCH ON
We have outlined our long-term approach to studying the genetic basis
of Ascaris infection in an isolated human population from eastern Nepal.
The statistical genetic models that we have developed are dependent
upon the availability of
large pedigrees that cut across multiple
