Grateful to an outstanding multidisciplinary team spanning machine learning, clinical virology, and laboratory diagnostics across UWMedicine and Fred Hutch: Lucas J. Liu Youyi Fong Greg Pepper Aniela Sobel Alex Greninger Saraswathi Sathees Keith Jerome
Using a two-stage approach (detection + classification), the model achieved >98% accuracy compared to expert readers in a pilot dataset. Our next steps include incorporating indeterminate results, prospective clinical evaluation, and extending this approach to other assays.
In this study, we developed BlotDx, a deep learning framework that is a first step towards building a diagnostic assistant for image-based assays using the HSV Western Blot as a use case.
The HSV Western blot is the gold standard for serologic diagnosis, offering excellent sensitivity and specificity. However, it remains labor-intensive, requires expert interpretation, and is not widely accessible.
Happy to share our recent paper, now online in the Journal of Virological Methods: “BlotDx: A deep learning tool for Western blot-based diagnostics”
🔗 authors.elsevier.com/a/1moh5bm%7E...
www.sciencedirect.com/science/arti...
“[P]utting data into GISAID is like dropping it in a mail slot in an unmarked building,” says @colincarlson.bsky.social. “It’s wonderful that there’s so much cool stuff in that building. It would be great if we knew who owned it, or who paid for it, or what they plan to do with it.”
#IDsky 🧪
Also ironic that GISAID itself scrapes data from public databases like Genbank without proper attribution. Thanks for covering this.
On Oct 1, 2025, GISAID informed us that they had ended updates to the flat file of SARS-CoV-2 genomic sequences and associated metadata that we had used to update Nextstrain analyses since Feb 2020. GISAID's stated rationale was that their "resources are limited". 1/5
😍
Helen Chu is one of the ousted #ACIP members. She spoke to me today about her grave concerns for future vaccination policy and vaccine access in the United States. www.statnews.com/2025/06/10/c...
Amazing! Congrats!
Absolutely thrilled to announce the launch of ARTIC2 - a £5.5M 5 year project funded by @wellcometrust.bsky.social to build on the ARTIC approach of low-cost, globally accessible genome sequencing for surveillance of outbreaks, epidemics and endemic diseases: www.birmingham.ac.uk/news/2025/am...
Cover of Emerging Infectious Diseases journal. https://wwwnc.cdc.gov/eid/articles/issue/31/13/table-of-contents
The May 2025 supplemental issue of Emerging Infectious Diseases journal is now live: “Advances in Pathogen Genomics for Infectious Disease Surveillance, Control, and Prevention.”
wwwnc.cdc.gov/eid/articles...
1/ How has H5N1 affected this year’s flu season in Seattle? Despite ongoing outbreaks in birds & cattle nationally, we detected zero H5N1 cases in 675 influenza A-positive specimens tested at UW Medicine.
6/ Read more in our manuscript from the
Greninger Lab
at UW Virology
: medrxiv.org/cgi/content/.... Many people contributed to this work, including
@margaretgracemills.bsky.social, Ailyn Perez-Osorio, and
@pavitrarc.bsky.social.
I’m so tired of hearing Rubio & other rightwingers now claiming that if you’re on a green card you’re a ‘guest’ in this country. BS. You’re a lawful permanent resident. The clue is in the name. If someone is a permanent resident in your home, are they the same as the guests who come and go? Come on.
Grateful to my amazing colleagues and coauthors for contributing to this work. And our collaborators across the pond who generously shared early versions of the ARTIC primers. /end
tl;dr The assay shows robust performance across a range of viral loads allowing sensitive recovery for RSV genomes from clinical specimens. Its sensitivity makes it suitable for use in clinical trials of antivirals and vaccines allowing early detection of treatment emergent/resistance mutation. 4/
We determined assay sensitivity, specificity, repeatability and reproducibility; and comparison with Sanger, shotgun, and capture-based sequencing methods to check for accurate recovery of the F gene. We also used contrived samples to determine the LOD of minor variants in allelic mixtures. 3/
We modified 2 ARTIC primers based on initial tests with samples from USA (WA state), but changes are likely improve performance in global isolates based on current variation at those sites. We were able to recover high quality genomes from clinical samples with a wide range of Cts (95% LOD Ct 30) 2/
Happy to share our latest preprint describing clinical assay validation for RSV whole genome sequencing using a modified version of the ARTIC RSV panel doi.org/10.1101/2025... <thread> 1/
Excited to be involved in this study led by Helen Chu as part of the CDC's CHARM network tracking the spread of respiratory viruses across the US.
My brain as soon as it saw this pop up on my feed: "In the shuffling madness of the locomotive breath" 🚂
My dad introduced us to Jethro Tull, listening to these on audio cassettes in the car. Love every single song.
A group of school children sit in front of an orange truck which is a plow/sander. They are holding up a hand-drawn sign which I can't read but I think it says THANK YOU
New names for Vermont snow plows 2024, selected by Vermont schoolchildren. Some selections...
Skibidi Scooper
No More Mr. Ice Guy
The Snowy Pl-Owl
Clearopathra
Snoah Kahan
Salt-O-Saurus
vtrans.vermont.gov/name-a-plow
I think you probably meant this Evan Newell--I don't think he's on bsky.
www.fredhutch.org/en/faculty-l...
