Use of AI-Based Algorithms to Breed Adaptive Plant Species

The use of artificial intelligence (AI) in the agricultural sector is gradually gathering momentum. Despite its limited applicability in agriculture, AI holds tremendous potential for improvements within bioengineering and crop protection. Meanwhile, Dr. Jacobson from the Oak Ridge National Laboratory (ORNL), US has proposed an AI-based solution for increased agricultural productivity. Dr. Jacobson’s team has done extensive research on plants and their response to climatic conditions. They also studied the adaptability of these plants to changing environments and climates.

 Developing Algorithms for Genome Study

 In order to establish a relationship between plant responses and AI technologies, Jacobson used high-order computing technologies. He deployed the IBM AC922 Summit supercomputer and the Cray XK7 Titan to expedite his research. Both of these computing technologies are available at the Oak Ridge Leadership Computing Facility (OLCF). Jacobson, along with his team, developed a new technique known as “genomic selection”.

 The development of the technique involves AI and machine learning algorithms to study the response of plant genomes. This shall help the researchers in developing or ‘designing’ organisms for the purpose of breeding. The algorithm helps in determining the type of genome variations that need to be initiated to produce adaptive plants. In this way, scientists can breed plants that can adapt to changing conditions and climates.

 Combining Past Successes with Current Developments

 Last year, Jacobson won the Gordon Bell Prize for successfully developing a code for studying organismic interactions. The code helps in understanding how organisms interact or communicate with their immediate environment. This code was used for the new research, in addition to the AI-based genome selection algorithms. This strengthens possibility of development of environment selection into a wider area of research.