Robustness analysis of metabolic predictions in algal microbial communities based on different annotation pipelines

Animals, plants, and algae rely on symbiotic microorganisms for their development functioning. Genome sequencing genomic analyses of these provide opportunities to construct metabolic networks analyze the metabolism communities they constitute. Genome-scale network reconstructions rest information gained from genome annotation. As there are multiple annotation pipelines available, question arises what extent differences in impact outcomes analyses. Here, we compare five commonly used (Prokka, MaGe, IMG, DFAST, RAST) predicted features (coding sequences, Enzyme Commission numbers, hypothetical proteins) network-based analysis (biochemical reactions, producible compounds, selection minimal complementary bacterial communities). While Prokka IMG produced most extensive networks, RAST DFAST fewest false positives connected with dead-end metabolites. Our results underline between outputs tested at all examined levels, small draft resulting different microbial consortia expand capabilities algal host. However, generated yielded similar compounds could therefore be considered functionally interchangeable. This contrast selected community functions depending pipeline needs taken into consideration when interpreting complementarity In future, experimental validation bioinformatic predictions will likely crucial both evaluate refine coupled increased efforts improve annotations reference databases.