Nathan Grubaugh Profile picture
Aug 22, 2019 22 tweets 35 min read Read on X
🚨 We used travel surveillance and genomic epidemiology to detect an unreported #Zika outbreak at an unexpected time.

Felt more like writing a detective novel 🕵️‍♂️🔎 than a scientific article...

“Zika files: case of the missing outbreak”

👇thread

cell.com/cell/fulltext/…
Tl;dr ▶️ watch the video abstract 👀

Our story started in 2016 by following local and travel case reports wanting to know when the Zika epidemic would be over.

We found this interesting spike in travel-related Zika cases from the Caribbean in 2017, when local cases were waning 🧐
🔎👩‍💻Looking more closely, we found that most of these 2017 Zika infected travelers were coming from Cuba… but Cuba didn’t report any local cases (at least not to the international community)

A missing outbreak?!
Now it was time to investigate 🕵️‍♂️👩‍💻👨‍🔬.

We found that the epi curves from local incidence rates matched travel incidence rates (travel cases/total travelers) from other locations, giving us confidence that the travel cases from Cuba were indeed from a missing local Zika outbreak
We then constructed a model 📈 to estimate that 1,000-20,000 Zika cases should have been reported in Cuba from 2016-2017 (187 Zika cases were reported in 2016 & none were reported in 2017).

Our estimates would put this on par with reported Zika outbreaks in the Caribbean 🏝️
@TAlexPerkins @RJOidtman So we detected an unreported Zika outbreak in Cuba, but the second part of this mystery is why did it occur in 2017 instead of 2016, when Zika outbreaks were peaking throughout the Americas 🌎?

And should we have expected this delay? 🤔
@TAlexPerkins @RJOidtman Based on our ✈️ travel surveillance, the #chikungunya outbreak in Cuba (which also was not reported) coincided with the rest of the Caribbean.

So no, outbreaks in Cuba are not somehow just universally disconnected from the rest of the Caribbean/Americas 🌎.
@TAlexPerkins @RJOidtman Next we ask, was the Zika outbreak delayed due to low transmission in Cuba, or because ZIKV was established later?

Here’s where we turned to the sequencing data 🦠🧬, which shows ZIKV was established in Cuba one year after other locations.
@TAlexPerkins @RJOidtman @gkay92 @GuyBaele Now the question is what caused the delayed establishment of ZIKV in Cuba?

Was it due to less ✈️ travel from places 🌎 with Zika outbreaks? No🚫
Was it due to unsuitable mosquito 🦟 conditions for outbreaks? No🚫
Was it due to mosquito 🦟☠️ control? Maybe❓
@TAlexPerkins @RJOidtman @gkay92 @GuyBaele Cuba reported a large mosquito 🦟 control campaign starting in early 2016.

Its timing⌚️ corresponded with a drop 📉 in dengue cases and low ZIKV transmission, despite suitable predicted mosquito transmission potential 🌡️🦟.
@TAlexPerkins @RJOidtman @gkay92 @GuyBaele Thus it looks like Cuba did an excellent job initially at control...

but perhaps the program waned and there were still too many opportunities for ZIKV to become established - and 💥 outbreak. (last part speculation)

👇 🦟Spray truck in Havana (@NBCNews)
@TAlexPerkins @RJOidtman @gkay92 @GuyBaele @NBCNews Who cares if there was an unreported outbreak that happened at an unexpected time? 🤷‍♂️

Well, because the world is so connected ✈️🌏, this could lead to silent 🤫 transmission to other locations. #yourproblemsaremyproblems
@TAlexPerkins @RJOidtman @gkay92 @GuyBaele @NBCNews @MOUGK We believe that our framework of travel surveillance and genomics can help to solve other mysteries too 🔎.

And importantly, help to identify gaps in our disease surveillance and reporting systems.
@TAlexPerkins @RJOidtman @gkay92 @GuyBaele @NBCNews @MOUGK Oh behalf of our Sharada Saraf (badass undergrad!),
@gkay92, @profpathogen, Scott Michael & @K_G_Andersen

We’d like to say how awesome it was to work (again) with such a great team. Our collaboration with the Florida DOH -led by Andrea Morrison’s - has been especially great 👏
@TAlexPerkins @RJOidtman @gkay92 @GuyBaele @NBCNews @MOUGK @profpathogen @K_G_Andersen Also SO to Kamran Khan's team that provided the travel volume data and @TAlexPerkins, @RJOidtman, @TexasDownUnder, @GuyBaele, @VogelsChantal, @jtladner, & many many others for providing data, analysis, and helping to design the study.
@TAlexPerkins @RJOidtman @gkay92 @GuyBaele @NBCNews @MOUGK @profpathogen @K_G_Andersen @TexasDownUnder @VogelsChantal @jtladner And special SO to @profpathogen who did all of the artwork 👩‍🎨, including the cover of Cell ‼️

See her work 👉 steampunkphage.com
@TAlexPerkins @RJOidtman @gkay92 @GuyBaele @NBCNews @MOUGK @profpathogen @K_G_Andersen @TexasDownUnder @VogelsChantal @jtladner @sciencecohen @sciencemagazine @jennifergardy @MichaelWorobey article by @carlzimmer in the @nytimes points out that the 'missing' outbreak may be a data mixup between Cuba and PAHO... 🤔 needs some more digging.

If Cuba is reporting ~1300 cases from 2017, then it validates our work.

• • •

Missing some Tweet in this thread? You can try to force a refresh
 

Keep Current with Nathan Grubaugh

Nathan Grubaugh Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!

PDF

Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @NathanGrubaugh

Jul 15, 2022
⚡️We developed & performed initial validation on an amplicon-based sequencing protocol (PrimalSeq) for Monkeypox virus (MPXV).

It improves genome coverage compared to metagenomic sequencing, especially at CTs >20.

🧵 1/8 | Protocol 👉
protocols.io/view/monkeypox…
It's based on a design led by @Scalene & @pathogenomenick originally for Zika virus that was adapted for SARS-CoV-2 ("ARTIC protocol") and used by labs around the world.

Our goal was for this to be plug n' play with current SARS-CoV-2 protocols. (2/8)
nature.com/articles/nprot…
The primers were designed using PrimalScheme using a pre-outbreak A.1 clade reference genome (GenBank accession: MT903345).

The scheme comprises a total of 163 primer pairs with an amplicon length ranging between 1597 and 2497 bp (average length of 1977 bp). (3/8)
Read 8 tweets
Jul 14, 2022
🧬July-14 Connecticut #SARSCoV2 variant surveillance
@CovidCT|@CTDPH|@CDC_AMD|@YaleSPH|@Yalemed

📈BA.5 is now the dominant variant in CT (53% from 6/27-7/4), followed by BA.2 (34% & 📉) and BA.4 (11% & 📈)

🚫No detection of BA.2.75

🧵 (1/8) | Report 👉
kphamyale.shinyapps.io/variant_report/
Using a logistic regression of the daily frequencies, we predict that as of today (July-14), BA.5 is probably 80-90% in Connecticut.

BA.4 is still 📈 as it outcompetes BA.2, but will probably start to 📉 in frequency soon after BA.2 is gone. (2/8)
We created a new dashboard to report variant sequencing data in Connecticut. You can still access it through our main website by clicking on the "Read the latest Connecticut report" link. (3/8)

👉 covidtrackerct.com
Read 8 tweets
May 19, 2022
🧬May-19 Connecticut #SARSCoV2 variant surveillance
@CovidCT|@CTDPH|@CDC_AMD|@YaleSPH|@YaleMed

📊Sequencing last 3 weeks
➡️ BA.2.12.1 = 44% (📈 8%)
➡️ BA.2 (other lineages) = 56% (📉 7%)
➡️ BA.4/5 = 0.3% (5 total)

🧵 (1/13) | Report 👉
covidtrackerct.com/variant-survei… Image
Omicron BA.2.12.1 is still 📈 in Connecticut as it is across most of the US. Fitting the % of sequenced cases to a logistic growth curve, we estimate that BA.2.12.1:

1⃣ is ~80% frequency today (May019)
2⃣ surpassed 50% in early May
3⃣ may reach 95% in early/mid June
(2/13) Image
From the same logistic growth curve, we also estimate that BA.2.12.1 is:

➡️ ~24% more transmissible than background (mostly other BA.2 lineages)
➡️ doubling in proportion every ~12 days
(3/13) Image
Read 15 tweets
May 5, 2022
🧬May-5 Connecticut #SARSCoV2 variant surveillance
@CovidCT|@CTDPH|@CDC_AMD|@YaleSPH|@Yalemed

Last 3 weeks
➡️BA.1.1 = 4%
➡️BA.2.12.1 = 23% (may be slowing down)
➡️BA.2 (other lineages) = 73%
➡️BA.4/BA.5 = 0% (might still be coming)

🧵 (1/9) | Report 👉
covidtrackerct.com/variant-survei…
We no longer have TaqPath data from YNHH to track SGTF, but we will use our "VOC" PCR assay that detect SGTF and ORF1a target failures (ORFTF).

BA.1 = SGTF
BA.2 = ORFTF

BA.1 is barely hanging on. (2/9)
👆 here is the "VOC" PCR assay referenced above. (3/9)

journals.plos.org/plosbiology/ar…
Read 9 tweets
Mar 17, 2022
🧬 March-17 Connecticut #SARSCoV2 variant surveillance
@CovidCT | @CTDPH | @CDC_AMD | @YaleSPH | @Yalemed

📊Sequencing data last 3 weeks
BA.1/.1 = 84%📉
BA.2 = 16%📈
Delta = 0%

📊PCR data this week
BA.1/.1 = 41%📉
BA.2 = 59%📈

🧵 (1/7) | Report 👉
covidtrackerct.com/variant-survei…
Based on our TaqPath PCR data (S-gene detected), we estimate that:

➡️ BA.2 is >50% in Southern Connecticut
➡️ At this rate - BA.2 will be 95% by early April
➡️ BA.2 doubling rate = 7.8 days (BA.1 in December = 3-4 days)
➡️ BA.2 ~43% more transmissible than BA.1/.1

(2/7)
Over the past 4 weeks, all of the sequenced S-gene positive samples have been Omicron BA.2 and not Delta. So we trust the 👆 PCR results reflecting the rise in BA.2. (3/7)
Read 8 tweets
Feb 3, 2022
Here are comparative results between 10 TaqPath S-gene detected samples tested by YNHH and with our validated VOC PCR assay. Most with our assay were actually SGTF, and looking at the YNHH results, the S-gene CTs for those were 5-7 higher than N/ORF. (5/16) Image
We are looking into these low level spike amplification samples that should be SGTF to see if this is a lab/TaqPath assay artifact or if there is something about these BA.1 sequences. So far doesn't seem to be sequence-related. Will report (6/16)
Our initial SGTF case definition – ORF/N <30 CT, S “not detected” - was conservative to not over-call BA.1.

We updated it yesterday to include S-gene 5 CTs higher than ORF/N, and compared the results. (7/16) Image
Read 13 tweets

Did Thread Reader help you today?

Support us! We are indie developers!


This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Don't want to be a Premium member but still want to support us?

Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal

Or Donate anonymously using crypto!

Ethereum

0xfe58350B80634f60Fa6Dc149a72b4DFbc17D341E copy

Bitcoin

3ATGMxNzCUFzxpMCHL5sWSt4DVtS8UqXpi copy

Thank you for your support!

Follow Us!

:(