Research

Project A4

Genetic correlations and the analysis of complex traits

Lead Partners 9+B: NPZ+JLU (SME + University) 1: QMUL (University)

State-of-the-Art Problem and its Solution
Next Generation Sequencing (NGS) can characterise both the genome and the expression profile of the same individuals, giving the opportunity to correlate these data to the performance for complex traits like crop yield or seed oil content. Performance prediction models based on multivariate analysis of genome-wide SNPs and global transcriptome profiles can potentially accelerate crop breeding progress by enabling genomics-assisted selection of the best cross combinations. Covariation among SNP frequencies and multivariate expression profiles may reveal key sites of interest for selection, but could alternatively be spurious correlations due to the demographic history. It is therefore necessary to distinguish these historical effects from the patterns of commercial interest, in order to generate prediction models that are applicable in both closely related breeding populations and more distant gene pools. In this project, two ESRs will be located at Justus Liebig University in Giessen, Germany and Queen Mary, University of London, UK, respectively. Together the ESRs will will generate and analyse next-generation genomic data from a commercial oilseed rape (canola) breeding programme of NPZ-Lembke (Hohenlieth, Germany), developing statistical tools for increasing seed yield and quality in one of the world's most important oilseed crops.
Objectives
  • Generation, bioinformatic processing and statistical analysis of genome-wide SNP data and global expression profiles in genetically diverse lines from a commercial oilseed rape breeding programme.
  • Extension of models for neutral variation among populations to characterise the influence of multivariate genomic background variation among lines, and use of this distribution to analyse co-variation between SNPs and gene expression.
  • Validation of the models using only parental genome and transcriptome data in elite hybrid combinations, with no prior knowledge about their performance.

Methodology
Transcriptome data and SNP profiles are being collected in ongoing research work by JLU in genetically diverse, elite oilseed rape breeding lines. Additional genomic data will be acquired during the course of this project as one component of the ESR training.

Genomic estimated breeding values for genome-wide SNPs will be calculated for complex phenotypic traits, including seed yield and seed quality parameters, by comparisons with multi-year, multi-environment field trial data provided by the commercial partner. Co-variation among breeding populations will be modeled as the result of a genealogical process constrained by unknown ancestral population sizes and known relationships among lines.

Candidate SNPs will be identified as outliers from the null distribution of SNP-expression co-variance generated by this process, and the performance of this approach will be compared with a full Bayesian model fitting an effect size for a subset of loci. Effectiveness will be assessed by comparing the phenotypic performance (yield and seed quality) of related and non-related breeding lines with the predicted performance, based on the genomic selection models. The best-performing models will be implemented within the practical breeding programme of the commercial partner.

ESRs training by research
The ESR based at JLU will be involved in acquisition and management of high-throughput SNP and sequencing-based transcriptome data, and will be involved in bioinformatic and statistical analysis of large genotypic and phenotypic datasets.

The QMUL-based ESR will develop Bayesian modeling methods for performance prediction, using multivariate approaches to combine different dimensions of data on the genome and transcriptome level. Together the two ESRs will collaborate closely on the statistical interpretation of the data and validation of the prediction models in a commercial crop breeding context.

The ESRs will have opportunities to cooperate with Projects A3 and A6 in the implementation of genomics data for genome-wide polymorphism detection and quantitative trait analysis. Furthermore they will gain insight into commercial plant breeding and seed production facilities, along with opportunities to interact with other crop genomics research projects involving high-throughput next-generation genome data analysis. Both ESRs will take part in bilateral research exchanges between the partners in Germany and the UK. During these exchanges they will receive multidisciplinary training in diverse applied aspects of next-generation genomics, bioinformatics, biostatistics, quantitative genetics and practical plant breeding in a commercial context.

Job Description 6
Job Description 7