National Barley Cluster: Activity 6 – CropSNP: Ultra-low-cost genotyping in barley and soybean

Objectives:

• Assemble a set of lines representative of the following crops: 500 soybean cultivars and breeding lines (maturity groups (MG) 000-1) and 500 barley cultivars and breeding lines.
• Perform dense genotyping on these two sets of germplasm using genotyping by sequencing (GBS) technology to establish an extensive catalog of single nucleotide polymorphism (SNP) markers that are informative for each crop.
• Perform SNP genotyping on segregating populations for the two crops to produce a high-resolution consensus genetic map providing the genetic distance between SNP markers.
• Select “background” markers conferring uniform coverage and trait-specific SNP markers linked to traits of interest.
• Develop a crop-specific genotyping tool capable of yielding information on this selected subset of markers for less than $2 a sample.
• Perform genotyping using these crop-specific tools on the same sets of lines originally used (in step 1) to confirm the reliability and overall performance of these tools.

Impacts:

• The genotyping tool that was developed through this work provides soy and barley breeding programs with the DNA markers they need to develop better varieties at a price point that remains affordable.
• This technology has been largely adopted in many of the ensuing Cluster-funded breeding programs (2023-2028) and will enable these breeders to incorporate more and better genomic information in their work.

Scientific Summary:

Modern breeding programs rely on the use of molecular markers to facilitate the speed and efficiency with which improved varieties are developed. Two main types of markers are used: gene- or trait-specific markers and background markers. The former are used to select breeding lines that have a specific trait or quality (e.g., resistance to a disease) while the latter are needed to help breeders select parents and progeny with the best overall genetic makeup. Typically, these two distinct needs have been met using different technological platforms and the cost of testing gene-/trait-specific markers increased with each additional marker. In this work, we aimed to provide breeders with an all-in-one genotyping tool: thousands of background markers and dozens of gene-/trait-specific markers at a low and fixed cost.

Results:

The first two objectives aimed at providing breeders with a rich dataset of DNA markers on relevant breeding materials. The participating breeders thus provided lines that they felt were most relevant to their work and best captured the genetic diversity within their breeding programs. For each set of 500 lines (one for barley and one for soybean), we performed SNP genotyping using our established protocols. Each breeder was provided with both program-specific marker data and data for the entire set of genotyped lines. For barley, the complete data set comprised 618 lines (294 from Western Canada and 324 from Eastern Canada) and close to 80,000 SNP markers while, for example, the Brandon data set included 145 barley lines and 51,000 informative markers. These data were provided to each breeder in February 2020 and marked the completion of objectives 1&2.

The third objective centered on the production of genetic maps based on GBS-derived SNP markers, as the available maps published at that time were mainly for SNP markers obtained using SNP arrays (a different and more costly genotyping tool). Such genetic maps provide breeders with a better understanding of how markers are expected to behave within the breeding populations that are at the heart of the breeding endeavour. Two parallel efforts were carried out in soybean and barley, resulting in successive publications in 2021 and 2022. In the first case, over 3,700 barley lines were genotyped at a total of over 50,000 markers, and this allowed us to produce a very high-quality consensus genetic map. This map provides extensive and complete coverage of the barley chromosomes and is devoid of any “gaps”, as had been an issue with previous maps. Similarly, in soybean, over 1,800 lines were genotyped at over 16,000 markers to produce a consensus genetic map providing extensive and uniform genome coverage.

In the fourth objective, we had two related aims: to develop an optimized genotyping procedure yielding a lower, but sufficient number of background markers and to develop a few dozen SNP markers that were of greatest significance and interest to breeders. This was done by designing an entire novel approach to simultaneously provide data for these two different types of markers. As described in the two papers below (de Ronne et al. 2023 and de Ronne et al. 2024), an innovative library-construction procedure was designed to allow the integration of the two marker types (background and gene-/trait-specific) into a single library that, upon sequencing, would yield all the desired information. In barley, we identified a set of 34 gene-/trait-specific markers based on the input of breeders and provide data on up to 6,000 background markers.   

In objective 5, we designed and tested these two, crop-specific genotyping tools targeting SNP markers of interest to breeders. In brief, a multiplex PCR reaction to interrogate a set of 27 gene-/trait-specific SNP markers for soybean was designed and tested (AmpSeq approach). Similarly, we successfully designed and tested a multiplex to interrogate 34 barley markers. Most of these were related to disease resistance, flowering/maturity or spike development. When testing this multiplex on a set of barley lines, we obtained a mean of 76 reads per SNP per line, with a minimum coverage of 2 reads and a maximum of 136 reads. In principle, for fixed lines a single read would be enough to accurately call the genotype at a SNP marker. In our test data, 100% of the cases met this requirement. For segregating lines, where some loci might be heterozygous, we would ideally want a minimum of 5 reads to confidently call the genotype. In 97% of cases among our test dataset, at least 5 reads were obtained. In summary, for the PCR multiplex for simultaneously interrogating 34 SNP markers of interest to barley breeders, we can successfully obtain the desired result in more than 97% of cases.

The sixth and final objective of the project was to perform a validation of the accuracy of our genotyping platform. We had initially proposed to repeat the genotyping of the lines we had originally characterized at the beginning of the project (Objective 1). Since then, genome assemblies for AAC Synergy and other barley cultivars have become available. So, instead of testing the rate of concordance between SNP calls made using our original GBS protocol and the new miniaturized NanoGBS, we opted to compare the genotype calls obtained via NanoGBS and the AmpSeq approach with the genotype of the corresponding base position in the reference genomes. Thus, the thousands of background markers (from NanoGBS) and the dozens of trait-specific markers (from AmpSeq) can be directly compared to the “truth set” represented by the reference genomes. Although we obtained the necessary sets of data for a panel of 18 barley lines (including 6 for which a reference genome is available), we have yet to finalize the assessment of the accuracy of our genotype calls. Based on the preliminary data (small scale, visual inspection), however, we are confident that accuracy will exceed 95%. We expect these data to be submitted for publication in the coming months.

External Funding Partners:

Funding for the National Barley Cluster was provided by the Agriculture and Agri-Food Canada AgriScience Program through the Canadian Agricultural Partnership (2018-2023), with industry support from provincial barley and soybean farmer associations across Canada and SeCan Association.

Project Related Publications:

Abed, A., Badea, A., Beattie, A., Khanal, R., Tucker, J., Belzile, F. 2021. A high-resolution consensus linkage map for barley based on GBS-derived genotypes. Genome. 65:83–94.

de Ronne, M., Abed, A., Légaré, G., Laroche, J., Boucher St-Amour, V.T., Fortier, E., Beattie, A., Badea, A., Khanal, R., O’Donoughue, L., Rajcan, I., Belzile, F., Boyle, B., Torkamaneh, D. 2024. Integrating targeted genetic markers to genotyping-by-sequencing for an ultimate genotyping tool. Theoretical and Applied Genetics. 137:247.

de Ronne, M., Légaré, G., Belzile, F., Boyle, B., Torkamaneh, D. 2023. 3D-GBS: A universal genotyping-by-sequencing approach for genomic selection and other high-throughput low-cost applications in species with small to medium-sized genomes. Plant Methods. (19):13.

Fallah, M., Jean, M., Boucher St-Amour, V.T., O’Donoughue, L., Belzile, F. 2022. The construction of a high-density consensus genetic map for soybean based on SNP markers derived from genotyping-by-sequencing (GBS). Genome. 65: 413–425.