Mr Gabriel Maicas1, Thomas Mooney, Lisa Chandler, Kate Montgomery, Andrew Harford, Joshua Koh
1Department Of Climate Change, Energy, The Environment And Water, , Australia
Biography:
Gabriel Maicas is an AI Research Scientist with over 14 years of experience developing solutions across health and environmental science. His work focuses on applying artificial intelligence and data-driven approaches to address complex scientific and societal challanges.
Abstract:
Environmental DNA metabarcoding is increasingly used in routine biomonitoring, but analysis remains a bottleneck: the main tools require bioinformatics expertise and offer limited support for project management, quality control and traceability – features that monitoring programs in highly regulated sectors demand.
We present QIIMERA (pronounced “Chimera”), a Python-based bioinformatics toolbox built on Qiime2, alongside with other established R and Python libraries, with a graphical user interface (GUI) for end-to-end DNA amplicon analysis. The workflow proceeds through guided stages, from raw demultiplexed sequenced reads to primer removal, denoising, taxonomic assignment, phylogenetic tree building and rarefaction analysis. Each stage offers rich visualisations and data logging for traceability, with quality control reports for auditing purposes. In addition, experiments are conveniently organised into projects with shared assets (datasets, classifiers, recipes etc.), allowing for rapid prototyping and resource sharing amongst team members. A database manager is also provided for a centralised management of in-house and externally curated reference databases, including ad hoc retrieval of sequences from public repositories (NCBI, BOLD) and support for custom classifier training. We assessed QIIMERA against published studies for COI and 16S rRNA amplicon analyses, retrieving similar number of processed read count, assigned amplicon sequence variants and taxonomic composition as those reported
We expect QIIMERA to make DNA amplicon analysis accessible to both novice and expert users alike, with reproducibility and traceability core tenets by design – important features for any long-term DNA-based biomonitoring program.