September 24 (WNN) — Humans and plants have thousands of genes. Traditionally, studying the function of a single gene or group of genes has required extensive experimentation.
Computers and access to large databases of genomic data, however, allow researchers to study the functionality of genes more efficiently. Nevertheless, mining genomic data on a large scale is difficult for even the most powerful computers.
In a new breakthrough, researchers in the United States and Taiwan have developed a machine learning algorithm to more efficiently identify “genes of importance” in agriculture and medicine.
Machine learning algorithms, described Friday in the journal Nature Communications, could help scientists predict how plants and animals will respond to changes in nutrition, toxins or pathogens — helping researchers develop more resilient crops, fight rare diseases, and more. allowing to diagnose or predict the next epidemic. .
“We show that by focusing on genes whose expression patterns are evolutionarily conserved across species, our ability to learn and predict ‘genes of importance’ for developmental performance of staple crops as well as disease outcomes in animals.” increases,” senior study author Gloria Coruzzi, biology professor at New York University’s Center for Genomics and Systems Biology, said in a news release.
Essentially, the researchers found a way to reduce the genetic noise to which an algorithm is subject.
“We show that reducing our genomic input to genes, whose expression patterns are conserved within and across species, is a biologically theoretical way to reduce the dimensionality of genomic data, which can be used in our machine learning models. significantly improves the ability to identify which genes are important for a trait,” said lead author Chia-Yi Cheng, a researcher at the Center for Genomics and Systems Biology and National Taiwan University.
In a proof-of-concept experiment, the researchers showed that nitrogen-responsive genes are evolutionarily conserved between two diverse plant species: Arabidopsis, a small flowering plant and plant model popular among plant scientists, and several varieties of corn. With the noise reduced from the input data, the new algorithm efficiently and successfully identifies important genes for nitrogen processing.
Nitrogen absorption is essential for plant growth. Plants engineered to absorb and use nitrogen more efficiently can reduce fertilizer use. Nitrogen overuse has been implicated in a variety of environmental problems, including nutrient overloading, harmful algae blooms, and coral bleaching.
In follow-up experiments, the researchers confirmed the importance of the genes identified by their algorithm. Plant scientists were able to amplify the genes of corn varieties to increase nitrogen intake and improve plant growth in nitrogen-deficient soils.
Co-author Stephen Moose, professor of crop science at the University of Illinois at Urbana-Champaign, said, “Now that we can more accurately predict which corn hybrids are better at using nitrogen fertilizer in the field, we are increasingly able to characterize this.” can improve.” “Increasing nitrogen use efficiency in corn and other crops provides three major benefits by reducing farmer costs, reducing environmental pollution, and reducing greenhouse gas emissions from agriculture.”
In addition to identifying genes relevant to various crop traits, the researchers suggest that their algorithms could be used to anticipate genes relevant to disease outcomes in mouse models, which could lead to new treatments and diagnostic techniques. can stimulate development.
“Because we have shown that our evolutionary informed pipeline can also be applied in animals, it is important to uncover genes of importance for any physiological or clinical traits of interest in biology, agriculture or medicine. underscores the potential,” Coruzzi said.