Microbial communities are key drivers of biogeochemical processes in the ocean. Recent methods involving metatranscriptomic approaches provide powerful data to investigate the diversity and activity of microorganisms in situ. Despite this, interpretation of gene expression data in an ecological context is often constrained by a lack of reference genomes, standardised analytical practices, and reproducible workflows. These limitations hinder both comparability across studies and robust ecological interference. This thesis aims to address these challenges through the development and application of accessible, standardized, reproducible bioinformatic pipelines primarily for metatranscriptomic but also metagenomic analyses. nf-core/metatdenovo was developed as a reproducible workflow for de novo assembly, quantification, and annotation of metagenomic and metatranscriptomic data. This pipeline enables taxonomic and functional characterization without reliance on reference genomes, providing a practical solution for studies in under-characterized environments. To provide a suitable tool when reference genomes are available, I developed nf-core/magmap, a pipeline that maps metatranscriptomic reads directly to genomes–provided by the user or publicly available–allowing analysis of expression patterns at the level of individual populations. A comparative analysis of these two pipelines within this thesis shows how methodological choices influence functional annotation, taxonomic resolution, and, ultimately, ecological interpretation, highlighting the strengths and limitations inherent to each pipeline. Finally, I applied nf-core/magmap to a metatranscriptomic dataset collected across seasons along a coastal–offshore gradient in the northwest Iberian upwelling system to investigate microbial rhodopsin expression in nature. The results revealed spatial, seasonal, and lineage-specific patterns in rhodopsin transcription, with substantial functional heterogeneity among closely related taxa. Rhodopsin investment is shown to be associated with different metabolic strategies rather than a universal functional coupling, underscoring the ecological diversity of microbial light-harvesting strategies. Altogether, this thesis shows how reproducible approaches to the analysis of meta-omics datasets can be used in advancing the ecological understanding of microbial functions.
Danilo Di Leo (Thu,) studied this question.