New deep-learning framework Crop-GPA 2.0 enables transferable decoding of crop genotype-phenotype associations across species
Peer-Reviewed Publication
Updates every hour. Last Updated: 20-Oct-2026 05:16 ET (20-Oct-2026 09:16 GMT/UTC)
Building upon their earlier platform Crop-GPA 1.0, a team of researchers in China developed Crop-GPA 2.0 to enable transferable SNP-level genotype–phenotype association prediction across crop species. Across multiple crop species, Crop-GPA 2.0 improved cross-species prediction performance, remained robust under limited training data, and identified biologically meaningful SNPs supported by previously reported quantitative trait locus (QTL) evidence.
A Food-System Carbon Shift
Tropical islands support important terrestrial and coastal carbon stores while facing strong pressures from development, land conversion, tourism, and climate change. In Hainan Island, food production, aquaculture, processing, and associated supply-chain activities have become increasingly important sources of greenhouse gas emissions within a rapidly developing regional economy.
Authors Zichen Li, Zhaohai Bai, and colleagues from Hainan University, the Chinese Academy of Sciences, and Nanjing University reconstructed the carbon budget of Hainan's food system from 1952 to 2020. Their life cycle assessment framework incorporated offshore ecosystems, land-use and land-use change, crop cultivation, livestock, aquaculture, food processing and consumption, wastewater treatment, and on-farm and off-farm energy use.
Seven Decades of Rising Emissions
Gross food-system emissions increased from 3.29 Mt CO₂ eq in 1952 to 15.13 Mt CO₂ eq in 2020. The analysis attributes this rise principally to fertilizer-related emissions, expansion of tropical crop cultivation, aquaculture growth, and increased energy demand along the food value chain. Nitrogen fertilizer application increased from 223 kg ha⁻¹ to 627 kg ha⁻¹ over the historical period.
An international team of scientists featuring faculty at Binghamton University has recreated a 1,000-year timeline documenting the ecological changes and human impact to the area surrounding Crawford Lake, the famed location of the proposed Anthropocene.
An international team of scientists featuring faculty at Binghamton University has recreated a 1,000-year timeline documenting the ecological changes and human impact to the area surrounding Crawford Lake, the famed location of the proposed Anthropocene.
An efficient few-shot framework was developed for large-scale annotation of light-trapped insect images. MLTIDD integrated Grounding DINO, SAM, and SAHI for multi-scale insect detection, while InsectSSRL exploited massive unlabeled insect images to enhance few-shot classification. MLTIDD achieved 79.6% AP50–95 and 90.8% AR, while InsectSSRL achieved 80.35% classification accuracy using only 5% labeled data. The framework reduced expert annotation time by approximately 80%, enabling efficient large-scale dataset construction.
A new study published in Science Advances reveals how Bronze Age communities in Central Europe adapted to changing environmental and social conditions by adopting broomcorn millet, a fast-growing and climate-resilient crop. Using multiple lines of bioarchaeological evidence, researchers found that millet was already being consumed in southeastern Poland around 1590 BCE – more than a century earlier than previously estimated – and was initially adopted gradually by a small number of individuals. The findings suggest that millet cultivation enabled communities to expand agriculture into previously marginal areas with poorer soils and drier conditions, reshaping how people used the landscape and strengthening food security.