We caught up with Martin to get his insights on working with several international consortia and on the challenges he faced while working on the IWGSC RefSeq.

Aside from the IWGSC, you have been involved in several international plant genome sequencing projects. What would you say are the benefits of being part of an international consortium?

A great advantage of working in a large international consortium is that you can distribute the different tasks among experts in sequence assembly, genome mapping, gene annotation, gene expression and genetic diversity, etc. The whole process would have taken so much longer if a single or a few groups would have had to develop all the necessary skills on their own.

I have been part of several genome projects in cereal crops: bread wheat, emmer wheat, durum wheat, barley and rye. It has been proposed that the Triticeae should be considered “a single genetical system”. For the purpose of genome assembly, this is certainly true: the genomes have a similar structure and genetic maps are highly collinear and can be used of cross-validation of genome assemblies.

What did you find most challenging about your work on the IWGSC RefSeq project?

The most challenging aspect was developing efficient data management strategies to integrate the wealth of resources in wheat genomics: sequence assemblies, genetic maps, physical maps, sequence tags, Hi-C data. This was a crucial requirement to use all available resources for improving the initial whole genome assembly delivered by NRGene to the chromosome-scale pseudomolecules we have now.

Were you surprised by the quality of the hexaploid wheat whole genome assembly produced by NRGene? What would you have expected? Why is it different?

Two years ago, no one would have expected that it is possible to assemble accurate megabase-scale sequence scaffolds from Illumina data only. When I saw the statistics of the assembly that NRGene had constructed, it seemed too good to be true. But to our astonishment, all the very large scaffolds were collinear with our barley reference genome (which we had just finished by then) and later on, we were able to validate them with our Hi-C map of Chinese Spring.

Why is it so important to have a reference sequence of high quality?

The first thing that comes to mind is that gene discovery in biparental mapping population is so much easier now. Researcher can now easily put their genetic mapping results in the context of the physical genome to get a precise idea of how large their mapping interval is and how many genes are in there. And with some luck, their flanking markers land on the same sequence scaffolds and they might even be able to find their causal mutation with the help of the gene annotation.

Have you had an opportunity to assess how well the genomes meet the needs of plant scientists / plant breeders? If they still need improvement, what resources would you add?

I am involved in several map-based cloning/genetic mapping projects in barley, and for these the reference genome has really helped a lot. Also for transcriptomics, the gene annotation (in wheat and barley) is an indispensable reference. There are still some gaps in very complex regions (for example resistance loci). Closing these gaps will probably require long sequence reads, either at the whole-genome level or for selected BACs.

According to you, what are the next challenges for the application of genomics in wheat breeding?

The new genomic resources have given us the tools for finding the causal genes for Mendelian traits (e.g. induced morphological mutations) by genetic mapping. Understanding the molecular basis of more complex traits such as yield and baking quality (where there is a strong interaction with the agricultural environment) will not only need genome sequences, but also accurate and comprehensive datasets from many field trials as well as elaborate statistical methods to analyze them.

Publication date: 09/04/2017