Table 3 |
||||
|
Consistency between QTL detected with 'raw' and 'curated' genotypic data. |
||||
|
Total number of QTL detected (p < 0.05) |
||||
|
|
||||
|
Methoda |
Correlation between QTL profilesb |
Raw dataset |
Curated dataset |
Overlapc |
|
|
||||
|
SML |
0.895 ± 0.085 |
38 |
34 |
29 (81%) |
|
MR |
0.998 ± 0.002 |
86 |
84 |
57 (67%) |
|
CIM |
0.887 ± 0.128 |
83 |
86 |
57 (67%) |
|
|
||||
|
a SML, Statistical Machine Learning; MR, single Marker Regression; CIM, Composite Interval Mapping with 20 background markers at > 10 cM distance from the tested interval. b QTL profiles are whole-genome plots of the fraction of variance explained vs. genome position similar to those displayed in Figures 6-8. The values reported are means ± SD across the nine traits investigated in this study. cThe percentage overlap was computed by division by the average number of QTL detected with the two datasets. |
||||
|
Bedo et al. BMC Genetics 2008 9:35 doi:10.1186/1471-2156-9-35 |
||||
