Escrever um comentário

The emergence of H5N1 influenza viruses in U.S. dairy cattle is an important example of how viruses can move between species and adapt to a new host. This paper investigates how and when H5N1 avian influenza spilled over from wild birds into U.S. dairy cattle, focusing on two variants, B3.13 and D1.1, and how the virus has been evolving since that spillover.

The main research question is: when and how did H5N1, specifically the B3.13 and D1.1 variants, spill over from wild birds into U.S. dairy cattle, and how has the virus evolved after establishing itself in cattle? The study design is a genomic epidemiology study using whole-genome sequencing data from publicly available databases.

The main findings are that B3.13 spilled over into cattle once in late October 2023, while D1.1 spilled over twice independently in 2024 and may have circulated undetected for 5 to 6 months. The authors also found that both genotypes evolved faster in cattle than in birds, mainly because of relaxed purifying selection rather than increased positive selection, and site-specific analyses identified cattle-specific positively selected sites in HA.

The results do support the authors’ conclusions because they used multiple appropriate methods and arrived at consistent results. One limitation of the study is that the D1.1 cattle dataset is small, with only 43 samples.

A fair and ethical peer review should be done by scientists with strong expertise in the same or a closely related field and without conflicts of interest. One potential bias to watch for is that in a small subfield, reviewers may be able to infer the authors’ identities, and if the author is well established, this may make reviewers less critical.

Two strengths of the paper are that it uses a multi-segment concatenated approach that provides targeted and reliable output, and that its conclusions are supported by three different frameworks: RELAX, robust counting, and FUBAR. One major concern is that the conclusion of relaxed purifying selection in cattle for D1.1 is not strongly supported, with a p-value around 0.5, which is especially concerning given the small dataset.

Competing interests

The author declares that they have no competing interests.

Use of Artificial Intelligence (AI)

The author declares that they did not use generative AI to come up with new ideas for their review.