I want to thank everyone involved in this huge effort:
@broadinstitute.org @broadclinicallabs.bsky.social @pardissabeti.bsky.social @lakrasil.bsky.social @dannyjpark.bsky.social @massdph.bsky.social @sabetilab.bsky.social and so many others 🤗

This study illustrates that the combination of genomic and demographic data can yield a much more detailed understanding of the complexity of transmission patterns as they unfold across different geographic scales than either type of data alone

While we recognize the scale of SARS-CoV-2 genomic surveillance is no longer feasible, sequencing 500/wk was enough to detect emerging lineages and identify growth earliest. However, sequencing 50/wk was enough to detect lineages along the same timeline as we would have otherwise

While qualitative, we saw a sharp drop in infections among 5–11-year-olds after vaccines became available to that group. This decline—seen in both school and public testing—suggests substantial short-term protection from infection.

Across two genomic methods we found that more recent vaccination and booster doses reduced the likelihood of transmission. Boosted individuals ~35% less likely to initiate transmission than unvaccinated people.

In schools, colleges, and SNFs, we found much stronger evidence of within-facility transmission among student and resident-aged than among staff-aged. Staff-aged individuals had lower viral relatedness to their facility, suggesting more involvement in community transmission

We then examined within-facility transmission across colleges, schools, SNFs, and workplaces. Facilities overall had 2.7x more closely related virus pairs (within 2 mutations from each other) than community samples—evidence of sustained transmission in these settings.

Across three major Omicron lineages, new variants reached 50% frequency in college settings 4–13 days earlier than elsewhere. This early rise wasn’t observed in when colleges were out of session (during BA.5*), suggesting unique features of colleges that accelerate spread.

We leveraged demographic information (age + facility) in our dataset to explore how different settings shaped SARS-CoV-2 transmission. Colleges, in particular, emerged as early indicators of new variant spread—often detecting variant growth days before other sectors.

We also show that SARS-CoV-2 moved more quickly between metro areas than it did to the primarily rural areas between them. Overall, we find that viruses diffuse throughout the entire state in about 2 months

New variants became prevalent more quickly in the Boston suburbs than in other regional cities, despite more frequent external introductions into the latter—highlighting the role of local transmission from initial introductions into Boston in driving early variant spread.

Using high-resolution spatial and temporal data, we tracked the introduction and spread of six Omicron lineages across Massachusetts. Urban areas played a central role—both as entry points for new lineages and as key drivers of their dissemination.

Our analyses fall into 4 main categories:
1) How do viruses enter and move within MA?
2) What are the demographic drivers of transmission?
3) How did vaccination/boosting work against infection and transmission?
4) How much sequencing did we need to detect lineages?

The Broad Clinical Labs generated >130k SC2 genomes 🧬 between Nov 2021 and Jan 2023. We were able to match >85k individuals to metadata like age, sex, city of residence, facility of collection, and vaccination status