Fred just finished quality control of the IWGSC Whole Genome Assembly (WGA) and is currently working on the sequence annotation and the integration of the WGA data with all chromosome-based data generated by the IWGSC project leaders.

While convalescent after a running accident, Fred took the time to answer our questions and share his experience and views about the future of the IWGSC.

What did you find most challenging about the assembly and analysis of chromosome 3B?

The main challenge was to produce a single pseudomolecule based on the initial assembly of BAC pools. The fact that BACs were sequenced by pools was a key step in the 3BSEQ project as it allowed us to reduce the cost while keeping a high quality sequence. However, this strategy had a strong impact on the amount of bioinformatics work needed! It required us to establish dedicated tools and databases from scratch to deal with such a complex situation.

What was the biggest surprise in the chromosome 3B?

It was definitely the clear relationships between function and structure of the chromosome. While preparing the publication, I remember lab meetings where we realized how clear the chromosome partitioning was; whatever parameter we studied, we would fall into the same border delineating the chromosomal extremities as the dynamic compartment dedicated to adaptation. This was definitely unexpected. I felt we were truly making a discovery.

From your experience what does the IWGSC have to take most care of with delivering the sequence of the other chromosomes?

I am convinced that the sequence will be of high quality. The gene models will be used the most by breeders, rather than the sequence itself. Therefore, we have to really pay attention to the annotation because annotation mistakes could lead to misinterpretation, especially on the evolutionary aspects. The real challenge is to unify the annotation of the sequences to come up with an annotated reference sequence useful for breeders.

You are currently working on the IWGSC Whole Genome Assembly (WGA). Can you talk about the challenges and the aim of this work? How will it be integrated with other genomic resources generated by the IWGSC?

When NRGene provided to us the first assembly, the metrics look so great that no one could believe it! The first challenge was to validate the assembly by comparing it with all available genomics resources, such as maps and previous assemblies. This step was quite challenging because these resources would also contain errors, but we have now strong evidence supporting the high level of quality of the IWGSC WGA. I wish we could integrate as many resources as possible, because the sequence itself is of little interest to most of the wheat community; rather, the added value is in an integrated resource that combines annotations, maps, markers, and QTLs.

In your opinion, what will be the best approach to understand variation in wheat genomes?

Resequencing is of great value but mapping short reads to identify variations is quite limited by the repetitiveness of the genome. Having the possibility to align high quality assembled genomes is much more efficient, but it requires reaching high quality for every new accession. I am convinced that getting reference sequences for a set of diverse accessions would be of great value to study genomic variations in wheat.

What do you think the biggest advances in wheat research will be in the next 5-10 years?

Right now, there is a major boost with mutagenesis (CRISPR/Cas9), single-cell molecular analyses, and epigenomics. I think that the biggest advances will come from functional characterization of genes via high-throughput targeted mutagenesis. This year, we have already seen major publications on rust resistance genes that were identified by a combination of mutagenesis and genomics. These types of approaches will certainly increase in the near future.

How dependent will these advances be on having the genome sequence?

Absolutely critical. And the quality of this sequence and gene models will also be critical, especially because of the amount of redundancy typical in the wheat genome. In bread wheat, genes are mostly repeated, sometimes at a very high copy number, limiting considerably our ability, for instance, to easily transfer knowledge from model crops to wheat based on orthology/paralogy relationships.

Do you want to continue working on wheat, or move to other areas of research? Why?

I am comfortable with both. For someone interested in genome evolution, wheat and its wild relatives is a fantastic model.
I often think about moving to other areas because I am convinced this is the best way to innovate. I was a microbiologist, now I am a plant scientist. I often think about combining both, such as studying the impact of bacterial communities on wheat development. A fascinating topic.

Publication date: 05/18/2016