Our research focuses on understanding the ecology and functionality of soil microorganisms in an agricultural context. In this specific project, we investigated how the soil microbiome of a winter wheat field changes throughout the growth season: from fallow bulk soil, to the plant-influenced rhizosphere, and then back to fallow. To study the taxonomic and functional shifts in the microbial community over this 32-week period, we undertook extensive sequencing, in the hope of capturing rare taxa and specialised functional genes. This left us with a substantial amount of sequencing data and posed a significant computational challenge. While taxonomic profiling and functional annotation of the total reads could be completed using our own hardware, we were unable to attempt metagenome-assembled genome (MAG) assembly using the large read files, which were up to 80 GB in size per sample. Fortunately, we had learned about the de.NBI infrastructure and their metagenomics toolkit (Metagenomics-Toolkit) at a workshop hosted by TU Braunschweig. We therefore reached out to them to discuss the necessary virtual machine requirements for running these assemblies. The team provided valuable advice on the computational requirements and swiftly set up a VM for this specific task. For this experiment, we managed to assemble high-quality metagenome-assembled genomes (MAGs), which will complement the overall taxonomic and functional findings from this time series study. Early bulk soil samples yielded high-quality Actinomycetota metagenome-assembled genomes (MAGs) with large genomes and a broad enzyme arsenal for metabolising complex plant-derived carbohydrates. Later samples, which were more strongly affected by the wheat rhizosphere, enabled us to assemble Pseudomonadota MAGs that appeared to be better adapted for rapid growth in nutrient-rich microhabitats. The findings from this study have helped us to identify gene ensembles that describe the different evolutionary adaptations that benefit bacteria in plant-influenced soils versus those in which bacteria must exist independently of plant inputs. The results are currently under peer review for publication, and we are continuing to use the VM for this project. We are now mapping the raw reads back to our MAGs to characterise their occurrence patterns over the 32-week period, which will help us gain insight into the exact seasonal dynamics of these taxa. The ease of access and use, scalability, and computational power provided by de.NBI have allowed us to advance our research and compare different tools on a large scale. We would like to thank them for their excellent support and hope to continue working with them on this and other potential bioinformatics projects.
