Biology Reference
In-Depth Information
mapping
cloning approach for mutant alleles,
exploiting the power of SNP-mapping and next-generation
sequencing
[46]
, or using RNAi, will always be powerful
ways to open up new areas of biology. In addition to this,
the resolution at which we can view the organismal state is
ever increasing
rescue
global population. There are collections of recombinant
inbred lines (and advanced RIL collections with increased
mapping precision)
[155,186]
, introgression lines (ILs)
[188]
, and expression data for many of these which can
allow for eQTL mapping in a whole animal. Crucially,
these datasets and collections mean that we have an accu-
rate overview on the total genetic variation in the entire
C. elegans species and, given any specific quantitative trait
that varies between any two naturally occurring isolates, we
have the means to map the genetic variants that explain the
phenotypic variation. Thus, one might ultimately envision
a comprehensive analysis of the dissection of the genetic
basis for all phenotypic variation in C. elegans (or in
a related species such as C. briggsae)
e
e
from cellular level anatomy, we are
already entering the world of high-resolution molecular
anatomy
[184]
, any cell type being defined not only by
lineage, function, and morphology, but also by a genome-
scale view of molecular expression markers, and soon
metabolite markers are likely to follow. This increase in
resolution will feed into computational modeling as out-
lined above, and our view of the wild-type N2 animal and
how it got to be that way will always be increasing in
resolution. However, we will focus on a small number of
areas of potential future progress that address central issues
of animal biology and genetics which can only now begin
to be addressed thanks to the current state of knowledge and
technology.
The first is the move outside the cozy, contained world
of the N2 reference and into the far murkier reality of
natural variation. Perhaps social scientists will simply see
the guiding hand of postmodern relativism at work (invi-
olate axiomatic frameworks shattered, a world of slippery
inconstancy), but ultimately this is a vital area of genetics in
which the worm can make a tremendous impact. A central
problem in human genetics is to understand the genetics
underlying natural variation in quantitative traits, made
particularly acute by the advent of personal sequencing. At
the heart of this is a major issue of statistical power: given
an unlimited number of individuals and crosses, one could
map all the possible variants that affect any traits. However,
there are only 7 billion humans on earth
e
the insights into the
genetics of natural variation in another animal, man, would
be huge. We anticipate that in the coming decade or two
attempts to understanding the genetic architecture of
natural variation in the worm, and the computational
predictive methods developed, will play a major role in
understanding natural variation in humans and in
increasing our ability to predict our phenotypes in health
and disease from our personal genome sequences.
To study natural variation in quantitative traits not only
requires knowledge of genetic variation, but also requires
the accurate measurement of quantitative traits and we
anticipate that this will also be a major area of growth.
While many of the phenotypes studied in classical genetic
screens have been characterized in a quantitative manner
e
e
for example the number of vulval precursor cells induced to
a vulval fate
[35]
, or the number of corpses in the anterior
pharynx of
the developing embryo (for example,
in
[189]
)
this quantitation has been almost entirely manual
to date. However, an increasing number of groups are
pushing ahead in the development of automated pheno-
typing platforms on a variety of fronts. Analysis of static
images
[154,190,191]
, much like the high-content high-
throughput style of mammalian cell-based assays, is clearly
one option, and software such as CellProfiler
[192]
, an open
source image analysis package, may prove to be key tools
for the worm. Researchers have also tried to use automated
analysis of time-resolved image series, either analyzing
films of worms swimming or crawling
[193
e
an ecologically
massive figure, but for the examination of the key issue
regarding whether hundreds of relatively rare alleles of low
effect size may cumulatively explain a large component of
the variation in any trait
[185]
, this is not vast by any
means. Every human is unique, and a human geneticist
ultimately has to work at best with the 14 billion currently
existing human genomes. However, any worm can be
cloned indefinitely, and rational mating schemas can be
drawn up to generate collections of intercrossed lines which
allow the mapping of natural variant alleles that affect any
traits of interest
[155,186]
. Statistical power
e
197]
, or using
automated imaging for the automated lineaging of devel-
oping embryos
[129,130,198]
. Finally, broad population
phenotypes such as fitness have also been examined quan-
titatively using either a commercial worm sorter or even
food utilization rates
[101,199]
. The increased sensitivity
and automated statistical definition of mutant phenotypes
allows for far more complex genetic analyses than would be
possible for crude quantitative phenotypes. Finally, the
myriad possibilities opened up by microfluidics to trap,
sort, perturb, and image worms are also beginning to bear
fruit, and this will also clearly expand greatly as a field
in the coming years
[200
e
an unlimited
number of individuals and crosses, and even an unlimited
number of individuals of any genotype
e
is not an issue
(except in practice; we choose to ignore this confining issue
for now).
To begin to address natural variation in quantitative
traits, the worm is now well placed. Thanks to pioneering
efforts of the Felix and Kruglyak groups among others, in
addition to the reference N2 genome, there is an analog of
the human HapMap
[187]
delineating the genetic rela-
tionships among the major subgroups of the C. elegans
e
202]
. The well-defined body
e
