Skip to main content
Fig. 5 | Genome Medicine

Fig. 5

From: Ontology-aware deep learning enables ultrafast and interpretable source tracking among sub-million microbial community samples from hundreds of niches

Fig. 5

Successful source tracking of environmental samples from a less studied biome by using ONN4MST. Results were based on using 11 samples from groundwater environment, which represented a biome previously less studied. a–c Source tracking results by using ONN4MST at the second, third, and fourth layers. d Source tracking results by using FEAST. e Source tracking results by using SourceTracker. Actual biome of query sample: “Root-Environmental-Aquatic-Freshwater-Groundwater.” A_1, A_2: two samples collected from a single well; B_1, B_2: two samples collected from another single well; C_1, C_2: two samples collected from the third single well; D-H: samples collected from other five wells, respectively

Back to article page