Skip to main content

Table 1 Effect of experimental settings on polymorphic-marker selection and linkage-detection thresholds in the mPub BSA scan

From: A DArT platform for quantitative bulked segregant analysis

      

Linkage-detection threshold (p < 0.05)5

Bulk size

Number of replicate arrays

Marker-selection p threshold1

Minimum hybridization difference between parents2

Number of polymorphic markers identified3

Markers previously mapped in Steptoe/Morex population4

Based on 'platform noise'6

Based on 'pooling noise'7

20

4

0.001

0.88

515

257

26%

50%

20

4

0.0001

1.04

433

231

24%

50%

20

4

0.00001

1.18

384

211

23%

50%

20

2

0.0001

1.06

356

187

24%

50%

20

8

0.0001

0.66

669

294

24%

50%

40

4

0.0001

1.04

418

221

23%

37.5%

  1. 1A normal distribution-based threshold for log2 [cy3/cy5] derived from the comparison of two identical aliquots of a 1:1 mixture of the Steptoe and Morex parents.
  2. 2log2 [cy3/cy5]
  3. 3Markers were selected from the set of 2,304 polymorphism-enriched clones (see section entitled 'DArT assays' in Materials and Methods).
  4. 4DArT markers were mapped on an array containing a partly overlapping set of markers.
  5. 5Values are based on the dispersion of the relative hybridization contrast (log2 [cy3/cy5] as a percentage of log2 [cy3/cy5] measured in the parental comparison) or the allele-frequency difference. There was a 1:1 correspondence between the two (Figure 1).
  6. 6This significance threshold reflects the variability inherent in the array-hybridization process. It was derived from the dispersion of the relative hybridization contrast in a 'self' comparison between two identical aliquots of 1:1 mixture of Steptoe and Morex (= ratio between log2 [cy3/cy5] in the self and the parental comparison). The resulting significance threshold was Bonferroni-adjusted for multiple comparisons.
  7. 7 This significance threshold reflects the chance that a non-zero allele-frequency difference may occur by chance as a result of the random assortment of chromosomes (and unlinked areas within chromosomes) in the pooling process. It was derived by simulating the pooling process (see section entitled 'Allele-frequency determination and simulation' in Materials and Methods and Figure 2).