Artificial intelligence helps predict the performance of sugar cane in the field

A Brazilian study published in Scientific Reports shows that artificial intelligence (AI) can be used to create efficient models for the genomic selection of sugar cane and forage grass varieties and predict their performance in the field based on their DNA.

In terms of accuracy compared to traditional breeding techniques, the proposed methodology improved predictive power by more than 50%. This is the first time that a highly efficient genomic selection method based on machine learning has been proposed for polyploid plants (where cells have more than two complete sets of chromosomes), including the grasses studied.

Machine learning is a branch of AI and computer science that deals with statistics and optimization with myriad applications. His main goal is to create algorithms that automatically extract patterns from data sets. It can be used to predict a plant’s performance, including its resistance to biotic stresses such as pests and diseases caused by insects, nematodes, fungi or bacteria and/or abiotic stresses such as cold, drought or salinity or tolerant or insufficient soil nutrients.

Crossbreeding is the most widely used technique in traditional breeding programs. “You build populations by crossing interesting plants. For example, in sugar cane, you cross a variety that produces a lot of sugar with a more resilient strain,” says computer scientist Alexandre Hild Aono, first author of the paper on the study. Aono is a researcher at State University’s Center for Molecular Biology and Genetic Engineering (CBMEG-UNICAMP). of Campinas, graduated from the Federal University of São Paulo (UNIFESP).

“But this evaluation process takes a long time and is very expensive. The method we propose can predict the performance of these plants even before they grow. We managed to predict the yield based on the genetic material. This is important because it saves many years of evaluation,” explained Aono.

In the case of sugar cane, the challenge is very complex. According to Anete Pereira de Souza, Professor of Plant Genetics at UNICAMP’s Department of Biology and Ph.D. Supervisor at CBMEG.

“When breeders identify an interesting plant, they clone it so that the genetic material is not lost, but this takes time and a lot of money. An extreme example is growing rubber trees, which can take up to 30 years,” Suza said. She calls one way of overcoming these difficulties “plant breeding 4.0”, which makes intensive use of data analysis and highly efficient calculation and statistical tools. Each genotyping-by-sequencing process can include 1 billion sequences.

The biggest hurdle scientists face when trying to breed better varieties of polyploid plants like sugar cane and forage grass is the complexity of their genomes. “In this case, given the scarce resources and the difficulty of working with this level of complexity, we didn’t even know if genomic selection would be possible,” Aono said.

methods

The researchers began the genomic selection process with diploid plants (which contain cells with two sets of chromosomes) because they have simpler genomes. “The problem is that high-value tropical plants like sugarcane are polyploid rather than diploid, which is a complication,” Souza said.

While humans and almost all animals are diploid, cane can have up to 12 copies of each chromosome. Each individual of the species Homo sapiens can have up to two variants of each gene, one from the father and the other from the mother. Sugar cane is more complex because theoretically each gene can have many variants in the same individual. There are regions of its genome with six sets of chromosomes, others with eight, ten, and even 12 sets. “The genetics are so complex that breeders work with cane as if it were diploid,” Souza said.

In 2001, Theodorus Meuwissen, a Dutch scientist who is currently Professor of Animal Husbandry and Genetics at the Norwegian University of Life Sciences (NMBU), proposed genomic selection to extract complex traits in animals and plants associated with their phenotypes (observable traits, which result) predict from the interaction of their genotypes with the environment). The advantage of this approach to plant breeding is the linkage between the phenotypic traits of interest such as yield, sugar content or precocity and single nucleotide polymorphisms (SNPs). A “snip” (as SNP is pronounced) is a genomic variant at a single base position in DNA, Souza explained.

“It’s the difference in the genome of any two individuals. For example, one may have an A (corresponding to the nucleotide adenine) that produces slightly more than another with a G (guanine) at the same location in the genome. That’s all changing,” she said. “If you find an association with what you’re looking for, such as For example, if you have high sugar production and specific SNPs at different locations in the genome, you can only sequence the population that your breeding work is focused on.”

The advances proposed by Aono and colleagues eliminate the need for planting and phenotyping throughout the breeding cycle. “We conduct field experiments in the initial phase of the program to obtain the phenotype of interest for each clone,” Souza said.

“At the same time, we sequence all clones of the breeding population in a very uncomplicated way, without having to have the entire genome for each clone. This is called genotyping-by-sequencing – partial sequencing looking for differences and similarities in the base pairs for the clones and their association with the production of each clone. The association between phenotype and genome shows which produces more and which SNPs are associated with higher production. This allows us to identify clones with a large proportion of the SNPs that contribute to the higher production observed in the initial experiments and to obtain the highest-yielding variety more quickly and cheaply.”