|
|
Falsely Declaring a Marker-QTL Association
Case 1
Lines
Lines
Locus 1 2 3 4 5 6
1
AA AA
AA aa
aa aa
2
BB BB
BB bb
bb bb
3
cc
cc cc
cc cc
cc
4
MM MM
MM mm
mm mm
Mean 6/3=2 0/3=0
Case 2
Lines
Lines
Locus 1 2 3 4 5 6
1
AA AA
AA aa
aa aa
2
BB BB BB
bb bb
bb
3 CC
CC CC
cc cc cc
4 MM
MM MM
mm mm
mm
Mean 9/3=3
0/3=0
In Case
1, a Type I error would occur because the lines with the
MM marker genotype have a mean of 2 and the lines with
the mm marker genotype have a mean of zero. Locus 3 is
linked to the marker, but there is no segregation for
the QTL in that population. In Case 2, locus 3 is segregating
and the lines with the MM marker genotype have the CC
genotype for the QTL. Lines 4, 5, and 6 have the mm marker
genotype and the cc genotype for the QTL. A correct decision
is made in rejection of Ho: no marker QTL association
or no QTL present for Case 2.
Case 3
Lines
Lines
Locus
1 2 3 4 5 6
1
AA AA aa
aa AA aa
2 bb
bb
BB BB
bb BB
3 CC
CC CC
cc cc
cc
4
MM MM MM
mm mm
mm
Mean 6/3=2
3/3=1
In Case 3, the correct
decision is made to reject Ho and declare the presence
of a marker-QTL association. All the lines with the MM
marker genotype have the CC QTL genotype and all the lines
with the mm marker genotype have the cc QTL genotype.
The background genes average out to 4 positive alleles
for each set of lines. This is what would be expected
in most cases.
Copyright
2000©, Ted Helms
|
|
