SE146:/DS2

Sample Set Information

ID	TSE1303
Title	Exploring molecular backgrounds of quality traits in rice by predictive models based on high-coverage metabolomics
Description	BACKGROUND: Increasing awareness of limitations to natural resources has set high expectations for plant science to deliver efficient crops with increased yields, improved stress tolerance, and tailored composition. Collections of representative varieties are a valuable resource for compiling broad breeding germplasms that can satisfy these diverse needs. RESULTS: Here we show that the untargeted high-coverage metabolomic characterization of such core collections is a powerful approach for studying the molecular backgrounds of quality traits and for constructing predictive metabolome-trait models. We profiled the metabolic composition of kernels from field-grown plants of the rice diversity research set using 4 complementary analytical platforms. We found that the metabolite profiles were correlated with both the overall population structure and fine-grained genetic diversity. Multivariate regression analysis showed that 10 of the 17 studied quality traits could be predicted from the metabolic composition independently of the population structure. Furthermore, the model of amylose ratio could be validated using external varieties grown in an independent experiment. CONCLUSIONS: Our results demonstrate the utility of metabolomics for linking traits with quantitative molecular data. This opens up new opportunities for trait prediction and construction of tailored germplasms to support modern plant breeding.
Authors	Redestig H, Kusano M, Ebana K, Kobayashi M, Oikawa A, Okazaki Y, Matsuda F, Arita M, Fujita N, Saito K
Reference	BMC Syst Biol. 2011 Oct 28;5:176.
Comment

PubMed

Data Analysis Details Information

ID	DS2
Title	Data processing (LC-MS)
Description	The profiling data files recorded in the MassLynx format (raw) were converted to the NetCDF format using the DataBridge function of MassLynx 4.1. From the set of NetCDF data files, the data matrix was generated using the MetAlign software (De Vos et al., 2007). By using this procedure, the data matrixes with unit mass data were generated. The data matrices were processed using in-house software written in Perl/Tk. The original peak intensity values were divided with that of the internal standards (lidocaine at m/z 235 [M + H]+ and (–)-camphor-10-sulfonic acid at m/z 231 [M–H]– for the positive and negative ion modes, respectively) determined in the same samples to normalize the peak intensity values among the metabolic profile data.
Comment_of_details