Crop researchers harness artificial intelligence to breed crops for the changing climate

Beans, Carolyn

doi:10.1073/pnas.2018732117

Cited by 22 publications

(15 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The conceptual framework we have introduced and our results show that adopting multi-5 6 7 8 9 objective optimization tools from operations research to solve breeding problems is highly advantageous (Cameron et al 2017;Beans 2020;Kusmec et al 2021). Several improvements or extensions suggested next should also be considered.…”

Section: Discussionmentioning

confidence: 92%

“…For example, in a single cross of two genetically distinct parental lines with 100 QTLs governing all the desirable traits, assuming independent assortment and co-dominance, there are in total 3 100 ≈ 5.1 × 10 47 genotypic combinations in the progeny. Even considering a moderate number of 200 wheat lines in the parental stocks, the number of combinations to be evaluated in the Ąeld is astronomically high (Beans 2020). Consequently, analytical approaches based on operations research, mathematical optimization, and statistical learning to optimize breeding decision-making have been suggested in several recent works (Han et al 2017;Kusmec et al 2021).…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Optimizing expected cross value for genetic introgression

Chen,

Ahadi,

Balasundaram

et al. 2024

Preprint

View full text Add to dashboard Cite

In this study, we address the mate selection problem in the hybridization stage of a breeding pipeline, which constitutes the multi-objective breeding goal key to the performance of a variety development program. The solution framework we formulate seeks to ensure that individuals with the most desirable genomic characteristics are selected to cross, in order to maximize the likelihood of the inheritance of desirable genetic materials to the progeny. Unlike approaches that use phenotypic values for parental selection and evaluate individuals separately, we use a criterion that relies on the genetic architecture of target traits and evaluates the combination of the genomic information of the pair of individuals. We introduce the expected cross value (ECV) criterion that measures the expected number of desirable alleles for a gamete produced by two individuals of the population selected as parents. We use the ECV criterion to develop an integer linear programming formulation for the parental selection problem. The formulation is capable of controlling the inbreeding level between selected parents. We extend the approach in two directions: (i) improving multiple target traits simultaneously, and (ii) finding a multi-parental solution to design crossing blocks. We evaluate the performance of the ECV criterion using a simulation study. Finally, we discuss how the ECV criterion and the proposed integer linear programming techniques can be applied to improve the efficiency of genetic introgression while maintaining genetic diversity in a breeding program.

show abstract

Section: Discussionmentioning

confidence: 92%

Section: Introductionmentioning

confidence: 99%

Optimizing expected cross value for genetic introgression

Chen,

Ahadi,

Balasundaram

et al. 2024

Preprint

View full text Add to dashboard Cite

show abstract

“…In particular, the use of artificial intelligence approaches has the potential to extract extremely useful information from breeding data (Harfouche et al., 2019), but the complexity of its deployment and the need for high computational capacity would be better served by a centralized approach with commercial providers rather than by each breeding programs individually. New solutions are being designed to facilitate the interpretation of the results like the Operations Research Approach, a technology originally developed to optimize decision in industry, that is, now being adopted to facilitate breeding decisions based on big data analysis (Beans, 2020). Hence, it is probably not farfetch to consider that the next decades will see the establishment of true “one stop shop” for genomic breeding, offering solutions from genotyping to data storage all the way to analysis and decisions.…”

Section: What Might Workmentioning

confidence: 99%

What plant breeding may (and may not) look like in 2050?

Bassi¹,

Sánchez-García²,

Ortíz

2023

The Plant Genome

View full text Add to dashboard Cite

At the turn of 2000 many authors envisioned future plant breeding. Twenty years after, which of those authors’ visions became reality or not, and which ones may become so in the years to come. After two decades of debates, climate change is a “certainty,” food systems shifted from maximizing farm production to reducing environmental impact, and hopes placed into GMOs are mitigated by their low appreciation by consumers. We revise herein how plant breeding may raise or reduce genetic gains based on the breeder's equation. “Accuracy of Selection” has significantly improved by many experimental‐scale field and laboratory implements, but also by vulgarizing statistical models, and integrating DNA markers into selection. Pre‐breeding has really promoted the increase of useful “Genetic Variance.” Shortening “Recycling Time” has seen great progression, to the point that achieving a denominator equal to “1” is becoming a possibility. Maintaining high “Selection Intensity” remains the biggest challenge, since adding any technology results in a higher cost per progeny, despite the steady reduction in cost per datapoint. Furthermore, the concepts of variety and seed enterprise might change with the advent of cheaper genomic tools to monitor their use and the promotion of participatory or citizen science. The technological and societal changes influence the new generation of plant breeders, moving them further away from field work, emphasizing instead the use of genomic‐based selection methods relying on big data. We envisage what skills plant breeders of tomorrow might need to address challenges, and whether their time in the field may dwindle.

show abstract

“…Figure 2 shows the companies in the food industry, such as TOMRA sorting food (tomatoes), are among the few that take the advantage of artificial intelligence to develop machines that significantly improve the sorting of food [13].…”

Section: Sortingmentioning

confidence: 99%

Modelling Techniques to Improve the Quality of Food Using Artificial Intelligence

Sahni¹,

Srivastava²,

Khan³

2021

Journal of Food Quality

View full text Add to dashboard Cite

Artificial intelligence (AI), or AI/machine vision, is assuming an overwhelming part in the realm of food handling and quality affirmation. As indicated by Mordor Intelligence, AI in the food and refreshments market is required to enlist a CAGR of 28.64%, during the conjecture time frame 2018–2023. Artificial intelligence makes it workable for PCs to gain as a matter of fact, investigate information from the two data sources and yields, and perform most human assignments with an improved level of accuracy and proficiency. Here is a concise gander at how AI is expanding sanitation and quality activities. This exploration has along these lines tried to furnish policymakers with a way to assess new and existing strategies, while likewise offering a reasonable premise through which food chains orders can be made stronger through the thought of the executive’s practices and strategy choices. This survey centers on the AI applications according to four mainstays of food security that is food accessibility, food availability, food use, and strength.

show abstract

Crop researchers harness artificial intelligence to breed crops for the changing climate

Cited by 22 publications

References 8 publications

Optimizing expected cross value for genetic introgression

Optimizing expected cross value for genetic introgression

What plant breeding may (and may not) look like in 2050?

Modelling Techniques to Improve the Quality of Food Using Artificial Intelligence

Contact Info

Product

Resources

About