AI to Predict Plant Traits and Accelerate Crop Breeding
Scientists at the Timiryazev Academy have completed the first stage of developing a digital predictive breeding platform powered by artificial intelligence. The system is designed to significantly shorten the time required to develop new crop varieties and improve the precision of agricultural genetics.
New Climate, New Hybrids
In the 21st century, global agriculture requires an increasing number of new crop varieties, placing higher demands on plant breeders. Farmers need higher-yielding cultivars with improved characteristics such as enhanced nutritional value and synchronized ripening periods to support industrial-scale harvesting.
Climate change presents an additional challenge. Hybrids must tolerate higher temperatures and prolonged drought conditions. Shifting climate patterns are also introducing new plant diseases and increasing pest pressure, requiring varieties resilient to emerging threats. At the same time, modern large-scale farming requires crops optimized for mechanized cultivation and harvesting.
Oksana Lut, Minister of Agriculture of the Russian Federation, has set a strategic target for breeders: by 2030, Russia must achieve 75% self-sufficiency in seeds for all major crops. The effectiveness of breeding programs and their alignment with market requirements are critical to meeting that objective, the minister explained.
“Over the past few years, farmers have repeatedly faced return frosts and extended droughts. Traditional varieties are no longer sufficient, so science, together with business, is developing new solutions. Another challenge I have mentioned is rising production costs. As a result, demand is growing for varieties with higher yields and lower cultivation costs,” the minister emphasized.
Digital Breeding Platform
The new AI-based system developed by the Timiryazev Academy processes large volumes of genomic and phenotypic data faster and more accurately than traditional statistical approaches. Researchers obtain detailed insights into plant inheritance patterns, including key traits such as yield, fruit quality, disease resistance and environmental adaptability.
To train the neural network, the academy creates genetically diverse crop populations and evaluates them under varying agroclimatic conditions. Scientists then analyze the relationships between DNA structure and observable traits. This enables predictive breeding – eliminating weaker specimens before field planting and focusing on promising candidates. The technology also forecasts optimal crossbreeding combinations and identifies the agroclimatic conditions under which a new hybrid will deliver the highest yields.
The platform is currently applied to rapeseed and cabbage hybrids, with plans to expand to additional crops. By integrating machine learning analytics and predictive modeling, the system is expected to reduce the breeding cycle from six years to approximately two. Accuracy levels may reach up to 99%, increasing the overall efficiency and intensity of selection programs.
Having successfully completed the first stage and validated the results, researchers have moved on to developing AI models for complex, multi-factor plant traits.
Expanding Food Production
The completed platform is expected to be fully operational by 2027. The project is part of the Priority-2030 program, which aims to establish more than 100 modern universities in Russia as centers of scientific, technological and socio-economic development by 2030. Agricultural universities across the country are building research hubs focused on smart farming, digital technologies and services for agribusiness, genetics, biotechnology and accelerated breeding.
The Timiryazev Academy platform is positioned to catalyze startups and research initiatives focused on AI-driven agricultural data analysis. It will provide a technological foundation for national digital breeding programs and for developing crop varieties adapted to Russia’s diverse climatic conditions. Agritech companies and research organizations will be able to forecast the performance and viability of new projects using the platform.
These advances are expected to increase domestic food production and strengthen Russia’s position in global agricultural markets. The predictive breeding platform could also attract interest from countries seeking to rapidly expand their own seed production capabilities.
