Hyperbolic Topological Data Analysis Mapper Reveals Hidden Plant Growth Trajectories
· newswise.com · popularity 14/100 · 0 similar stories today · Mathematics & Statistics (Science)
Read the original at newswise.com →
Summary
Modern plant phenotyping produces complex, high-dimensional data that traditional tools struggle to interpret. Researchers developed the Hyperbolic Topological Data Analysis Mapper (HTDA-Mapper), an algorithm that embeds phenotypic data in Poincare ball space to preserve hierarchical, nonlinear and temporal relationships. The pipeline handles quantitative descriptors and raw images, using unsupervised contrastive learning to compare images without manual labels. Applied to more than 27,000 Arabidopsis thaliana seedlings grown under different nutrients and priming compounds, HTDA-Mapper resolved compound-specific effects, concentration-dependent responses and dynamic growth trajectories. It offers a geometry-aware way to turn high-throughput phenotyping data into interpretable biological insight.
Found in: Newswise. Rating basis: single outlet; outlet tier 2 (unlisted); 1 feeds; feed position 24.