Flu infection is a team effort, and clinic data misses part of the picture

Infection takes a crowd

The textbook picture of a virus infecting a cell is a lone particle doing the job alone. A preprint on influenza A argues that picture is wrong for flu. The influenza A particle-to-PFU ratio is approximately 400, indicating that only a tiny fraction of particles possesses independent infectivity, so most particles in a sample cannot start an infection by themselves. Productive infection is predominantly driven by genome complementation among incomplete particles under high-pMOI conditions, which means particles that each carry only part of the genome combine their pieces inside the same cell to make a working infection. m:…9252

The dose response fits that story. Influenza infection rate follows a sigmoidal curve and reaches saturation at a pMOI of 5 to 10, so infection climbs steeply as more particles arrive per cell and then levels off. The functionality of genomic segments within an influenza particle follows a stochastic process, resulting in a robust probability distribution for the number of functional segments per particle, which is a formal way of saying that which segments work in any given particle is a matter of chance, and the population as a whole covers the gaps. m:…9252

The practical consequence is about how transmission risk is assessed. If infection is a collective process, then counting infectious units as if each particle acted alone will misjudge how a real exposure behaves. A cloud of mostly incomplete particles can still be infectious when enough of them reach the same cells, and a risk model built on single-virion thinking will underrate that. m:…9252

Clinics see a different virus

A second preprint asks whether the samples we sequence even represent what is circulating, comparing clinical and community surveillance for SARS-CoV-2, RSV and influenza in Washington State. SARS-CoV-2 clade proportions differ significantly between clinical and community surveillance samples, so the mix of variants looks different depending on where the swabs come from. Spike mutations associated with monoclonal antibody resistance are more prevalent in community-based SARS-CoV-2 samples compared to clinical samples, which means the mutations that matter for antibody treatments are easier to miss if only clinic data is watched. m:…3304

There is also a measurement gap between the two settings. Clinical samples for SARS-CoV-2, RSV, and influenza consistently exhibit lower Ct values than community-based samples, which reflects higher viral loads in people who seek care than in people sampled in the community. That is expected, but it is also a bias: surveillance built only on the sickest people will see a different epidemic from the one actually spreading, and the preprint argues for a hybrid approach that keeps community sampling in the frame. m:…3304

Telling lung infections apart

A third preprint turns from counting viruses to reading the immune response, comparing COVID-19 with active tuberculosis. COVID-19 neutrophils exhibit reduced CD10 and CD16 expression compared to tuberculosis, indicating a shift toward immature, emergency granulopoiesis, while tuberculosis neutrophils exhibit higher maturation, receptor, and oxidative response gene expression compared to COVID-19. In other words, the two infections leave different fingerprints on the same cell type: one floods the blood with young, unfinished neutrophils, and the other shows a more mature, receptor-ready response. m:…5516

The distinguishing signal held up across datasets. Maturation-associated transcription is the most consistently reproduced axis distinguishing COVID-19 from tuberculosis across independent cohorts, which makes it a candidate biomarker axis for telling the two infections apart when symptoms alone cannot. That is useful in clinics where both diseases circulate and a fast, reliable distinction changes treatment, and it is a reminder that the immune system often knows which pathogen it is fighting before the tests come back. m:…5516