Progress 09/01/14 to 08/31/17
Outputs
Target Audience:Potato researchers, growers, and processors Changes/Problems:We were unable to complete Objective 2 due to lower than expected population size in the national trial and issues with marker data quality. Objective 2 was therefore refocused onsolving the marker quality problem,as described in the Accomplishments section. What opportunities for training and professional development has the project provided?This project has supported the training of a graduate student in plant breeding, with a specialization in quantitative genetics, as well asfunds topresent results from this project at scientific conferences. How have the results been disseminated to communities of interest?The results have been presented to other potato researchers at a number of scientific conferences (listed under Products). In addition,the PD has delivered extension-oriented talks covering results from this research at meetings for growers, processors, and other industry stakeholders. What do you plan to do during the next reporting period to accomplish the goals?
Nothing Reported
Impacts
What was accomplished under these goals?
Impact: The overarching goal of this project was to enable the use of molecular markers for potato breeding, recognizing that genetic improvement is a powerful approach to improving yield, processing quality, disease resistance, and other traits affecting the sustainability of potato production. Molecular breeding in potato has lagged behind other major crops in part because it is a complex autotetraploid, which means that two complete sets of chromosomes are inherited from each parent, rather than one set as in most species. Through this project, two new software packages were developed to better utilize tetraploid markers: one package generates high quality genotype calls from microarray data, and the other combines marker and trait information to identify causal genes or regulatory regions. These tools are now being used by the University of Wisconsin potato breeding program and other research groups around the world. Objective 1. Develop and test methods for genome-wide association studies (GWAS) in autotetraploid potato Simulation experiments were conducted using previously published marker data for 187 tetraploid potatoes, with the goal of evaluating the performance of different mixed models for genome-wide association analysis. In diploids, the primary technique used to speed up GWAS is known as the P3D approximation, in which the variance components from the null model are used in the significance test rather than re-estimating them for each marker. The first experiment determined that the P3D approximation had negligible impact on statistical power and false positive rate (FPR). The second experiment showed that the ability to test for dominant associations improved statistical power compared to the additive model when gene action at the causal locus is dominant. The implication of this result was that unique software was needed for polyploid GWAS, as existing diploid programs can only be readily used to test for additive associations in polyploids. We therefore developed and released the R package GWASpoly, which is available on the PD's website (potatobreeding.cals.wisc.edu/software) and is already being used by several other research groups. Objective 2. Identify genomic regions for marker-assisted selection through GWAS of agronomic and processing traits, such as reducing sugars, asparagine, and acrylamide, from the National Fry Processing Trial. This objective was formulated to be synergistic with the activities of the National Fry Processing Trial (NFPT), a variety development project funded by a consortium of fry processing companies and managed by Potatoes USA (formerly the US Potato Board). Because fewer than expected clones were evaluated in the NFPT, the population size was too small to achieve this objective. In addition, we identified some issues with the quality of the tetraploid marker data, which traced back to limitations with the methods used to convert signal intensity from the SNP array into genotype calls. Therefore, in lieu of completing the original objective 2, the graduate student supported by this Hatch project developed the R package ClusterCall (available at potatobreeding.cals.wisc.edu/software) to improve the quality of the marker data. ClusterCall uses multiple F1 populations to calibrate the relationship between signal intensity and allele dosage for each marker. For three potato F1 populations genotyped with an 8K array, ClusterCall scored 5729 markers with at least 95% concordance, compared to 5325 markers with the benchmark software fitTetra. When applied to the same diversity panel analyzed in Objective 1, ClusterCall produced 5218 high quality SNPs--a huge advance over the 3521 SNPs previously available. ClusterCall has become the foundation of molecular potato breeding at UW-Madison, providing higher quality marker data for genetic fingerprinting, GWAS, and genomic selection. Objective 3. Improve early generation screening of potato breeding lines for fry processing quality. In 2014 and 2015, FY4 selections from the russet breeding program were grown in a randomized complete block design with two replicates at the UW Hancock Research Station. The total number of entries including the checks was 42. Tubers from each plot were sent to the USDA Potato Worksite in East Grand Forks, MN, and stored for 3-4 months before processing. Tubers were cut into ¼" strips, blanched, par-fried, and then shipped on dry-ice to UW-Madison for a final 2:45 minute fry at 350oF. The overall color of finished fries was measured with a Hunter Colorimeter, and 50 fries per sample were scored for sugar end and textural defects (limp units). The reliability (equal to entry-mean heritability) of overall fry color was similar for the two trials: 0.54 in 2014 and 0.59 in 2015. The reliability of percent sugar ends was 0.42 in 2014 and 0.74 in 2015, and the results for percent limp defects were 0.34 in 2014 and 0.19 in 2015. The conclusion is that overall color and sugar end defects can be reliably measured in the FY4 stage, but limp defects are not produced consistently enough to select on this trait. We have therefore adopted a fresh-cut, slab frying protocol to assess overall color and sugar ends, which is simpler than the frozen, par-fry process needed to assess texture.
Publications
- Type:
Conference Papers and Presentations
Status:
Published
Year Published:
2017
Citation:
Schmitz Carley C, Endelman JB (2017) Genome-wide selection accuracy in tetraploid potato F1 populations. Potato Association of America Annual Meeting, July 2426, Fargo, ND.
- Type:
Journal Articles
Status:
Published
Year Published:
2017
Citation:
Schmitz Carley CA, Coombs JJ, Douches DS, Bethke PC, Palta JP, Novy RG, Endelman JB (2017) Automated tetraploid genotype calling by hierarchical clustering. Theoretical & Applied Genetics 130:717726.
- Type:
Conference Papers and Presentations
Status:
Published
Year Published:
2017
Citation:
Schmitz Carley C, Endelman JB (2017) Genomic selection in tetraploid potato. Annual Meeting of the National Association of Plant Breeders, Aug. 89, Davis, CA.
|
Progress 10/01/15 to 09/30/16
Outputs
Target Audience:Potato researchers, growers, and processors Changes/Problems:
Nothing Reported
What opportunities for training and professional development has the project provided?The graduate student supported by this project gave an oral presentation at the 2016 Annual Meeting of the Potato Association of America, in Grand Rapids, MI. She also gave a poster presentation at the Fifth International Conference on Quantitative Genetics, in Madison, WI. How have the results been disseminated to communities of interest?Through several conference presentations and a journal publication (see Research Products). What do you plan to do during the next reporting period to accomplish the goals?Objective 1. The objective is completed. Objective 2. The manuscript describing ClusterCall will be published in a refereed journal. Genome-wide association studies using fry and chip processing germplasm will be conducted. Objective 3. The objective is completed.
Impacts
What was accomplished under these goals?
Impact: The main goal of this project is to enable the use ofmolecular markers for new potato variety development, recognizing that genetics is a powerful approachtoimprovingthe sustainability of potato production. Significant progress toward this goal was made in 2015-16through the development and release oftwosoftware packages, as well aspresentations and publications illustrating their use. Objective 1. This objective was largely completed in 2014-15. The only remaining activity for 2015-16was to complete publication of the manuscript describing the GWASpoly software, which was done (Rosyara et al. 2016). The software is available under the GNU Public License (potatobreeding.cals.wisc.edu/software) and is already being used by other research groups around the world. Objective 2. In the 2014-15 progress report, we described an unexpected challengeregarding the reliability of the SNP marker data, which needed to be addressed to ensure the validity of the GWAS results. Because commercial potato germplasm is autotetraploid, there are three possible heterozygous genotypes for bi-allelic SNPs; this complicates conversion of the raw signal intensity from the SNP array into reliable marker data. In 2015-16, we completed the development of a new software package for making tetraploid genotype assignments, namedClusterCall (potatobreeding.cals.wisc.edu/software), which produces more accurate results than the current benchmark. When ClusterCall was used to make new marker calls forthe dataset analyzed by Rosyara et al. (2016) from Objective 1, an additional QTL for vine maturity was discovered using the GWASpoly software. A manuscript describing ClusterCallwas submitted for publication, and presentations were made at several conferences. Objective 3. A field trial of 42 russet varieties was conducted at the UW Hancock Research Station in 2015, using a randomizedcomplete block design with two replicates.The entry list included 32 FY4 entries, 2 checks (Russet Burbank, Umatilla), and 8 more advanced selections from the UW breeding program. Plots were mechanically harvested and graded to measure total yieldand the size distribution.In October, samples from each plot were shipped to the USDA Potato Worksite in East Grand Forks, MN, and stored for threemonths before processing. Tubers were cut into 1⁄4" strips, blanched, par-fried, and then shipped on dry-ice to UW-Madison for a final 2:45 minute fry at 350 degreesF. For each sample, 50 fries were scored manually using the USDA 0-4 color scale, and brightness (L*) was measured with a Hunter Colorimeter. The fries were also scored forsugar ends, color variation, and texturaldefects. The reliability, or broad-sense heritability (h2), of the data was quantified on a 0-1 scale, which representsthe squaredcorrelation between the true and estimated genetic values for that environment. Of the two methods for evaluating overall color, the manual approach had a higher reliability: h2 = 0.80 for the proportion of Color 0 fries vs. h2 = 0.54 for Hunter brightness. The reliability of the percentsugar end defect resultswas 0.74, and h2 = 0.63 for color variation defects. Unfortunately, the textural data were not reliable, with h2 = 0.19 for the %limp defects and h2 = 0.18 for % hollow defects.
Publications
- Type:
Conference Papers and Presentations
Status:
Published
Year Published:
2016
Citation:
Schmitz Carley C, Palta J, Coombs J, Douches DS, Endelman JB (2016) Automated tetraploid genotype calling by hierarchical clustering. Potato Association of America Annual Meeting, July 31Aug 4, Grand Rapids, MI.
- Type:
Conference Papers and Presentations
Status:
Published
Year Published:
2016
Citation:
Schmitz Carley C, Palta J, Coombs J, Douches DS, Endelman JB (2016) GWAS of tetraploid potato with automated genotype calls. 5th International Conference on Quantitative Genetics, June 1217, Madison, WI.
- Type:
Journal Articles
Status:
Published
Year Published:
2016
Citation:
Rosyara UR, De Jong WS, Douches DS, Endelman JB (2016) Software for genome-wide association studies in autopolyploids and its application to potato. Plant Genome 9, doi:10.3835/plantgenome2015.08.0073
|
Progress 10/01/14 to 09/30/15
Outputs
Target Audience:Potato researchers, growers, and processors Changes/Problems:
Nothing Reported
What opportunities for training and professional development has the project provided?The graduate student supported by this project attended the 2015 annual meeting of the Potato Association of America, in Portland, Maine. How have the results been disseminated to communities of interest?Presentations at scientific conferences (see Research Products section). What do you plan to do during the next reporting period to accomplish the goals?Objective 1. We will performany necessary revisionsand complete the publication of this research. Objective 2. We willprepare and submit amanuscript describing the ClusterCall software. We will also complete the marker curation and genotype calling for the NFPT population and conduct initial genome-wide association studies. Results will be presented at the 2016 annual meeting of the Potato Association of America. Objective 3. The reliability of the fry-processing traits will be measured for a second year.
Impacts
What was accomplished under these goals?
Objective 1. The first software package designed to perform genome-wide association studies in autopolyploids, called GWASpoly, has been completed and is available on the PI's website: http://potatobreeding.cals.wisc.edu/software. A unique feature of GWASpoly is its ability to model polyploid gene action, including additive, simplex dominant, and duplex dominant effects. This software was rigorously tested using simulated data and with the SolCAP diversity panel, which contained 187 tetraploid potato varieties phenotyped for 13 morphological, agronomic, and storage properties. The results confirm our hypothesis that higher statistical power is achieved when the gene action assumed in the GWAS model matches gene action at the unobserved QTL. Consistent with previous biparental mapping studies in potato, we identified strong QTL for potato shape and eye depth that co-localized on chromosome 10. The remaining QTL we discovered were only marginally significant, most likely because the dataset was underpowered for GWAS: for simulated traits with a heritability of 0.3, the median genome-wide power was only 0.01. Our results indicate that both marker density and population size were limiting factors for GWAS with the SolCAP panel. A manuscript describing this research was submitted for publication. Objective 2. There are three components to this objective: (1) Curation and summary of the phenotypic data, (2) Genotyping the NFPT varieties and curation of the marker data, and (3) Combined analysis of the marker and phenotypic data with GWASpoly. Regarding activity #1, curation and analysis of the 2011-2013 phenotypic data was completed as part of a related project. Phenotypes for the 2014 NFPT were completed during the project period, as well as some results for 2015. Regarding activity #2, we successfully completed the DNA extraction and genotyping of the NFPT entries using an Infinium array with more than 12,000 SNP markers. Our initial plan was to use existing methods for making the genotype calls. However, we discovered that the existing methods were insufficient, in one case because it could not be extended to the 12K array (the SolCAP cluster boundaries) and in the other because of unacceptable error rates (with normal mixture models, e.g., fitTetra and mclust). We therefore developed a new method for assigning genotype calls based on hierarchical clustering, called ClusterCall. Preliminary results indicate that 50% more SNP markers on the array pass QC with ClusterCall than with fitTetra. Activity #3 has not begun yet. Objective 3. As reported previously, forty-two russet varieties and advanced breeding lines with fry processing potential were grown in a two-replicate trial at the Hancock Agricultural Research Station in 2014. In Oct. 2014, samples from each plot were shipped to the USDA Potato Worksite in East Grand Forks, MN, and stored for four months before processing. Tubers were cut into ¼" strips, blanched, par-fried, and then shipped on dry-ice to UW-Madison for a final 2:45 minute fry at 350oF. The overall color of finished fries was measured with a Hunter Colorimeter, and 50 fries per sample were scored for sugar end, color variation, and textural (limp) defects. This was the first time such measurements had been made in the UW breeding program. The reliability, or broad-sense heritability (h2), of the data was quantified on a 0-1 scale, equal to the expected correlation between the true and estimated genetic values for that environment. The reliability of the overall brightness (L*) was 0.59, which is high enough to be useful for selection. For comparison, the reliabilities for specific gravity and tuber glucose in this population (measured on raw tubers) were 0.78 and 0.71, respectively. The sugar end and limp fry defect frequencies had reliabilities of 0.42 and 0.34, respectively, which are low but potentially still useful for eliminating very poor varieties. The frequency of color variation defects had essentially no heritability (h2 = 0.06) and would not be useful for selection. The 2015 cohort of field year 4 russet breeding lines were grown this season in a two-replicate trial at the Hancock Station to enable a similar experiment in the next reporting period.
Publications
- Type:
Conference Papers and Presentations
Status:
Published
Year Published:
2015
Citation:
Schmitz, C, D Douches, J Endelman. 2015. Population structure of the National Chip Processing Trial. Potato Association of America Annual Meeting, July 1922, Portland, ME.
- Type:
Conference Papers and Presentations
Status:
Published
Year Published:
2015
Citation:
Rosyara, UR, and JB Endelman. 2015. Genome-wide association studies for autopolyploids. Plant and Animal Genome XXIII, Jan. 1014, San Diego, CA.
|
Progress 09/01/14 to 09/30/14
Outputs
Target Audience:
Nothing Reported
Changes/Problems:
Nothing Reported
What opportunities for training and professional development has the project provided? The graduate student research assistant supported by this project received mentoring in data analysis and potato harvest logistics from the PI. How have the results been disseminated to communities of interest?
Nothing Reported
What do you plan to do during the next reporting period to accomplish the goals? Objective 1. The development of the autotetraploid software for GWAS will be completed, and the package will be made available to the public under the GNU Public License. The results of this research will be presented at the Plant and Animal Genome Conference in January 2015, and a manuscript will be submitted for publication. Objective 2. We expect to receive the first installment of marker data for the National Fry Processing Trial in Spring 2015. The graduate student will analyze the quality of the data and integrate it with other datasets to better understand the population structure of elite U.S. potato germplasm. Results will be presented at the Potato Association of America Annual Meeting in July 2015. Objective 3. Post-harvest evaluation of fry processing quality will be conducted, including novel assessments of fry internal texture and within-tuber variation for specific gravity. The 2015 field trial wil be designed and planted. Results will be disseminated at UW Agricultural Research Station field days and at meetings of the Wisconsin Potato and Vegetable Growers Association.
Impacts
What was accomplished under these goals?
Objective 1: Software development for autotetraploid GWAS is underway and being tested on both simulated data and phenotypes collected during the USDA-NIFA SolCAP project. Objective 2: Not yet begun. Objective 3: Our 2014 field trial of potato breeding lines for fry processing was harvested at the UW Hancock Agricultural Research Station on Sept. 26 and graded Sept. 29-30.
Publications
|
|