Re: HYB: Punnett Square

Chad Schroter wrote:

:        I have noticed a change in your postings since about 1 or 2 mont=
hs
ago, =

:  if you cannot tell from below, all the end of line markers are printin=
g
as an =

:  "=3D" sign. Perhaps this "3D" effect is related. Did you change your e=
mail
program
:  etc., because in the past I did not get this affect on your postings.

I haven't changed my mail program, but I did change from digest to regula=
r
mode sometime after bloom season.  I thought that affected only the forma=
t
I received, not what I sent.   I'll send one copy of this to the list and=

another directly, to see if it makes any difference.

:  A question about Possibilities vs. =

:  Probabilities

Before I tackle your questions, I think I should expand a bit on the
sequence.  The objective is to go from considering the possibilities, to
computing the probabilities, to deciding how to make the odds work in our=

favor.

If a parent carries a gene, even though it is not expressed, it can pass =
it
on.  If it doesn't carry it, it can't.  The trick is figuring out which
cultivars carry how many doses of which genes.  Often, we can only say th=
at
it is possible a certain one has the desired gene -- we can't tell for su=
re
without raising and observing a sufficient number of offspring and
sometimes the required number is impractically large.

In tetraploids, as Tom explained, each gene has a 50/50 chance of being
transmitted to each offspring.  Computation of probabilities springs from=

that. For a random-mating population, if we know enough about the parent
population to estimate the frequency with which a gene appears in it, we
can compute "Probabilities" without knowing much about the individual
parent.   We rarely know enough to do this with iris.

In selective breeding,however, if we know which genes a cultivar has, we
don't care about the frequency in the overall population.  We can constru=
ct
the table of possibilities, compute the frequency of occurrence of each
one, and come up with the probability of getting the one we want.

In yesterday's example, assume the goal is to obtain breeding stock that
can be crossed to glaciatas and produce glaciatas in the next generation =
--
so the desired genotypes are those which have two or three doses of pla. =
 =


Most of the seedlings will be violet (33/36 or 11/12 of the cells).  Two =
of
those will have three doses of pla, ten will have two doses, seven will
have one dose, and three will have none.  The chance of a particular viol=
et
seedling having the one of the desired genotypes is therefore 5 in 33.  N=
ot
great, but good enough that if the best plant of the cross had a violet
flower I'd go ahead and test it.

A few, however, will be plicata (3/36 or 1/12).  Two will have three dose=
s
of pla, one will have three.  All can produce glaciatas if crossed with
one, but 2/3 will produce glaciatas and plicatas in a 1:3 ratio while 1/3=

will produce them in a 1:1 ratio.  =


:   What factors besides relative gene location on individual or the =

:  same chromosome determine the "probability" of a parent providing a
particular =

:  gene pair ? =

 =

We always assume that each chromosome has an equal chance of being passed=

on.  This is not true for wide crosses, but Tom started the discussion fo=
r
TBs and I've tried to stick to that type of tetraploid as far as possible=
=2E =

In a random-mating population, the critical factor is thus the frequency =
a
certain gene appears -- there's no way of knowing whether a particular
parent carries it.  With selective breeding, however, we focus on the
individual.  The likelihood it carries a certain gene is determined by it=
s
ancestry.  =


Although we rarely know the genotypes with the certainly of our theoretic=
al
example, we can made some educated guesses.   Tom showed how to use the
phenotype to determine which ones must be present and which ones could no=
t
be present -- and left the unknowns as x's.  For example, we can't tell h=
ow
many doses of the t-factor a red or yellow self may carry from just looki=
ng
at it but  if we know that one parent was a tangerine pink, or that the
candidate itself has produced pink offspring, we know it must have at lea=
st
two.   We can carry pedigree analysis even further, by going back to
ancestors, estimating their genotypes, and carrying the transmission
probabilities forward.  (If you're working with dominants, or widespread
recessives, this is overkill.  You don't even want to think about how man=
y
such charts I've drawn in quest of tangerine pink arilbreds....)

:  If I understood TWOI correctly certain genes could be carried on =

:  separate chromosomes and thus be easy to separate or combine, while
others may =

:  be located closely together on the same chromosome and be relatively
difficult =

:  to separate (i.e.. both genes would tend to be present together or
absent =

:  together)

That is correct.  Genes that are located on separate chromosomes segregat=
e
independently.  Those that are located closely together on the same
chromosome are said to be linked.  Linkage can be an advantage if you can=

find a cultivar with the desired alleles linked -- an obstacle if the
"wrong" ones are linked.  The phenomenon that breaks linkages is called
"cross-over".  It happens during meiosis, when members of a pair of
chromosomes may break and exchange portions.  =


That's a more complex matter, however, involving two different traits
rather than a single set of alleles, so it doesn't lend itself to a simpl=
e
Punnett square.  To compute these probablities, you'd have to compute
separate squares, then use the linkage factor to relate them.  Beyond the=

scope of this "course"....

As for other factors -- there are many that don't directly affect the
probability that a certain gene is present, but do affect how it is
expressed and therefore how we perceive its presence or absence.  If you'=
re
ready for this material, I'd recommend books on genetics, topics like
epistatic genes, gene series, and gene interaction.

Sharon McAllister
73372.1745@compuserve.com

we must remember that "gene" is a man-made concept, a convenient
analyitical tool.  In reality, there are =
Other Mailing lists | Author Index | Date Index | Subject Index | Thread Index