Présentation
Routine virological surveillance is crucial for monitoring viral epidemics, particularly for identifying concerning mutations, evaluating viral genomic diversity, and predicting future dominant clades. High-throughput sequencing and efficient bioinformatics methods have been pivotal in this regard. Since the COVID-19 pandemic, numerous data analysis workflows have been developed for use in surveillance laboratories, targeting both respiratory (e.g., SARS-CoV-2, Influenza, RSV) and non-respiratory viruses (e.g., Ebola, Zika). These workflows typically map sequencing reads to a reference genome, detect single nucleotide polymorphisms (SNPs), and generate consensus sequences. However, for influenza virus, the presence of Defective Interfering Particles (DIPs) complicates accurate consensus reconstruction, introducing errors in viral consensus genomes and potential misinterpretations. Our goal is to automatically identify DIPs resulting from large deletions and correct consensus sequences accordingly.