Criteria for QTL selection and assignment to map positions

        Our criteria for including barley QTL reports in our summary were (i) the use of QTL detection tools, as opposed to Mendelian classification of the phenotype, and (ii) publication in a peer-reviewed journal in either paper or electronic format. Additional sources of QTL data include the Barley Genetics Newsletter, the GrainGenes database, and published conference proceedings. However, due to the preliminary and/or summary nature of such reports, we did not include them in our summary. We used the QTL descriptor, e.g. “grain yield” employed in the original report. In the case of multiple reports on the same QTL, the reports are listed chronologically. We did not include review papers in the database that do not provide additional information beyond the original report. QTL validation reports, when available, are shown in a separate field, following the original report.
         QTL were assigned to the Bin Map, herein referred to as the “BM”, of Kleinhofs and Graner (2000) as follows: if markers flanking a QTL were also in the BM, we were able to unequivocally assign the QTL to a bin or region. Binned QTL were assigned to the physical map of Künzel at al. (2000) - referred to as the “PM” - as follows. First, we identified markers that occurred in both the BM and the Igri/Franka map, referred to as “IF”. The IF map was used in developing both the BM (Kleinhofs and Graner 2000) and the PM (Künzel et al. 2000). We used markers in common to the BM and IF map as our starting framework. Then, IF markers without BM positions were assigned to the latter using the proportional distances to common flanking markers in the IF and CM maps. This generated an adjusted "Linkage Bin Map" (LBM ). We then aligned the LBM and the PM to generate a "Physical Bin Map" (PBM). The LBM and PBM bins were assigned recombination values, following the nomenclature of Künzel et al. (2000), to generate a "Recombination Bin Map" (RBM). We then aligned the RBM with the PM. As shown in the Figures, this allows for approximations of the physical size of the bins to which QTL are assigned and an estimate of the degree of recombination in each bin. Finally, as shown in the Figures, QTL assigned to each bin in the LBM were drawn proportional to the bin size in the PBM.
         In cases where there was one marker in common to the QTL report and the BM, the QTL was binned on the basis of markers present in the consensus map of Qi et al. (1996) - referred to as the consensus map, - “CM” - that coincided with markers in the BM. When markers flanking the QTL were not present in the BM, but were present in the CM and could be positioned relative to BM markers, we also assigned the QTL to single or adjacent bins. In some cases, QTL were assigned to one of several possible bins, based on the inferred position of the QTL relative to distant flanking markers that occurred consistently in the QTL report, the BM and the CM. In other cases, there were no consistent flanking markers and QTL could only be assigned to chromosome arms. RFLP and SSR markers were particularly useful for map integration and assigning QTL to bins, whereas results for anonymous and high throughput markers, such as RAPDs and AFLPs, were particularly difficult to interpret.