Radiation Hybrid QTL Mapping of Tdes2 Involved in the First Meiotic Division of Wheat

Radiation hybrid QTL mapping of Tdes2 involved in the first meiotic division of wheat

F.M. Bassi, A. Kumar, Q. Zhang, E. Paux, E. Huttner, A. Kilian, R. Dizon, C. Feuillet, S.S. Xu, S.F. Kianian

Supplementary information

A total of 426 DArT markers and 115 PCR-based markers have been previously mapped on 92 radiation hybrid (RH1) lines to generate a saturated map of chromosome 3B (Kumar et al. 2012a). That map spanned 1,871.9 cR as defined by 202 unique loci and had a calculated mapping error as low as 0.05 % (Kumar et al. 2012a). In that map, DArT markers have been anchored to other PCR-based markers to guarantee the high map quality. Ideally, all these markers should have been used for screening the entire 3B-RH population. Unfortunately, the default DArT genotyping platform provides low cost per data point but also high total experimental cost. As a strategy to simultaneously reduce the cost per data point, but to also maintain the total experimental cost in check when screening large populations, a mini custom array of 490 clones was produced (Fig. S1).

Fig. S1. The devised strategy to keep the cost low for DArT genotyping by screening an increasing number of lines with decreasing number of selected markers. The top dashed square represents future research endeavors.

The mini array contained a subset of 70 DArT markers from the 426 previously mapped by Kumar et al. (2012a) selected to evenly span the whole chromosome 3B. To these were added additional 175 DArTs that had been previously mapped in other genetic mapping populations (Wenzl et al. 2010; Huang et al. 2012). The 245 markers were then blotted in duplicate on the array. The overall mapping and genotyping strategy is depicted in Fig. S2, where a large population of 1,000 RH1 lines was divided into two subsets (182 and 696 lines) that were genotyped at different depths.

Fig. S2. Overall strategy by genotyping a subset of radiation hybrid populations to maintain low experimental cost, screen a large population, and achieve high map quality.

The 70 DArT markers mapped by Kumar et al. (2012a) were maintained in their pre-determined map position. Two point LOD were then calculated for the 70 markers in their pre-determined map order and in their best possible map order. All markers that had LOD less than 10% lower than their best possible position and were then left in their pre-determined order. However, six DArT and one SSR marker showed more than 10% lower LOD in their pre-determined position and were then excluded from the frame-work map. The pre-ordered 63 marker loci were considered as the frame-map and used to anchor the remaining 77 markers. Explanations of the iterative frame-work mapping approach, its importance and of the possible explanations for changes in the map position of markers are provided by Kumar et al. (2012a and 2012b). In Fig. S3 are shown the map position of 70 of the 140 markers mapped that are shared between the map presented here and the work conducted by Kumar et al. (2012a). The seven markers that had LOD lower than 10% of the max LOD in their pre-ordered position can be clearly identified as diagonal lines intersecting other lines. Of these, three are major map relocations while the other four are just minor changes.

Fig. S3. Autograph comparison of the map produced by Kumar et al. (2012a) and the one presented here. Parallel lines indicate conservation of map position, while intercrossing lines are indicative of changes in the order of markers.