Tweet

Alexandre Bolze

18 Feb, 9 tweets, 5 min read

@my_helix

SARS-CoV-2 surveillance by @my_helix update

- New dashboard: you can check all lineages now
Random ex: B.1.1.64 in 4 states (note: I don't know anything about this variant).
public.tableau.com/profile/helix6…
- SGTF info up to Feb 15
- Seq info up to Jan 30

🧵 with more results

2/

Of the variants of concern, so far we only identified B.1.1.7 (666 times up to Jan 30). No B.1.351 and no P.1

We identified many B.1.429, a variant of interest. It represents about 20% of the sequences we do every day. But sampling is still biased for SGTF

3/

To assess fraction compared to non-SGTF sequenced, you can also get that info from the 2 files on Github with ALL of the data.

Check it out
github.com/myhelix/helix-…

- Info on >300,000 positive tests in last 2.5 months
- Info on 2,843 sequences (2/3 are random non-SGTF)

4/

Tip: if you want to quickly download the 2 files on your computer.
- Click on the file on home page (github.com/myhelix/helix-…).
- Then click on "Raw". see picture.
- Then on your Chrome toolbar: File/Save Page as

=> your turn to play.

@scottleibrand

5/
It is very rewarding to see many use this information.

Example: @scottleibrand analysis of what might happen in next few months

https://twitter.com/scottleibrand/status/1362330346763603969?s=20

@trvrb

6/
@trvrb also referenced @my_helix work in this great thread of what we might expect in next few months.

https://twitter.com/trvrb/status/1362438562449293312?s=20

@ScottGottliebMD

7/
To finish, the usual snapshot of % of SGTF in Florida

Feb 14, 5-day avg was: 16% of positive tests are SGTF.
And experimental evidence support that >90% of these are B.1.1.7

Overall cases going down, vaccination is up, B117 is up. We keep tracking.

@ScottGottliebMD

@genesareclever

8/
Credit to many Helix teams. In particular @genesareclever and William Lee for leading the effort.

See our preprint for all involved.

medrxiv.org/content/10.110…

9/
So we made it a little easier. There is now a page with the fraction of B.1.427 or B.1.429 compared to non-SGTF sequences.
public.tableau.com/profile/helix6…

- check US or by state.
- look at total non-SGTF seq per collection day

Jan 29, 18% of non-SGTF we sequenced were B.1.429

• • •

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

This Thread may be Removed Anytime!

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

More from @alexbolze

Alexandre Bolze

@alexbolze

15 Feb

@my_helix

SARS-CoV-2 surveillance data from @my_helix is here: github.com/myhelix/helix-…

Stats on
>300,000 positives from last 4 months
2,503 sequences (1/3 SGTF, 2/3 random non-SGTF)

🧵on vars of concerns & interest in 7 states

>2% of positives are B.1.1.7 in: CA, FL, GA, PA & TX

@my_helix

@my_helix tests (PCR tests) in all states & all available on Github.

Analysis here limited to states with >100 positives PER DAY (and with N gene Cq<27) in December & January.
7 states: CA, FL, GA, IN, NC, PA, TX.

AZ, MA, MI were just under cutoff.

@K_G_Andersen

And analysis is focused on variants of concerns
B.1.1.7
B.1.351 (not present)
P.1 (not present)

variants of interest
B.1.427
B.1.429

See outbreak.info/situation-repo… from @K_G_Andersen or covariants.org from @firefoxx66 for info on these variants.

Read 12 tweets

Alexandre Bolze

@alexbolze

8 Feb

Thread about our preprint on the study of emergence and increased prevalence of SARS-CoV-2 B.1.1.7 variant in the US.

medrxiv.org/content/10.110…

2/
First, my Conflict of Interest: I am an employee of Helix.

We do a lot of COVID testing (PCR tests) and also receive funding for the viral surveillance program.

See the other COIs from authors of this preprint.

@my_helix

3/
Methods are very similar to the ones used in the regular @my_helix updates with the dashboard.
public.tableau.com/profile/helix6…

The study is based on >500,000 positives PCR tests. We test in all 50 US states, but it is not homogeneous. See map 👇

Read 13 tweets

Alexandre Bolze

@alexbolze

5 Feb

@my_helix

Latest numbers from @my_helix SARS-CoV-2 surveillance.

Data on PCR test positives & SGTF (S-gene target failure) up to Feb 3.
Stats on sequences and B117 up to Jan 22.

Link: public.tableau.com/profile/helix6…

Thread on Florida and California

2/
For each graph, numbers are provided for each day and as 5-day moving average.

n SGTFs and n positives come from PCR tests -> based on large numbers.

n SGTFs seq & n B117 come from sequencing -> smaller numbers. => 5-day avg important for these to reduce noise

3/
Florida first
- latest avg (n SGTF) / (n positives): 8.6%
- latest avg (n B117) / (n SGTF seq): 78%

Extrapolating the seq data, estimate (likely under estimate) is: B117 represents 6.7% of positives

Trend 📈

Read 7 tweets

Alexandre Bolze

@alexbolze

4 Feb

@my_helix

In @my_helix SARS-CoV-2 dashboard, we show that in Massachusetts, ~5% of positives are SGTF (= have del69/70). It has been consistent since November.

BUT, they are not B117. So what are these SGTF?

Majority are B.1.375.

More details in 🧵

public.tableau.com/profile/helix6…

B.1.375 is a SARS-CoV-2 variant with the following defining mutations:

Gene: protein change
ORF1a: T1828A
ORF1b: E1264D
ORF3a: T151I
Spike: del69/70 -> this is why it is a SGTF
M: I48V

Some details about the variant in this paper: virological.org/t/detection-of…

As far as I know, this is not a variant of concern.

In our sequences of SGTF samples, we also see this variant a lot in Florida. Following the fraction B1.1.7/B.1.375 in Florida over time will be informative to assess whether B.1.375 has any impact on transmissibility

Read 4 tweets

Alexandre Bolze

@alexbolze

3 Feb

@my_helix

Update on @my_helix SARS-CoV-2 variant surveillance program.
Data from testing with n positives & SGTF (S-gene target failure)
Stats on sequences (SGTF & non-SGTF)
Stats on B117

Link: public.tableau.com/profile/helix6…

Thread with analyses in CA, FL, MA ...

2/
First, the dashboard is still focused on B117. Because B117 is more infectious & had large impact in many UK, Ireland, Denmark...

But, looking for other variants is also super important. covariants.org

=> we are increasing sequencing of random positives. see map

3/
Back to SGTF & B117 analyses. Knowing there is a bias in selection, we estimate % of positives that are B117 by looking at 2 metrics:
- (n SGTF) / (n positives)
- (n B117) / n SGTF sequenced)

And all numbers are from positives with a N-gene Cq < 27.

Read 9 tweets

Alexandre Bolze

@alexbolze

27 Jan

@my_helix

New data on the @my_helix SARS-CoV-2 dashboard, which is still focused on B117 variant.

Link: public.tableau.com/profile/helix6…

Thread with specific analysis in California & Florida.

2/ Methods

All numbers are based on positives with Ct<27 for N gene.

Tracking 2 metrics to estimate % of positives that are B117:
- (n SGTF) / (n positives)
- (n B117) / (n SGTF sequenced)

Latest sample collection date:
for SGTF: Jan 22
for sequencing: Jan 15

@EricTopol

3/ CA results

Latest 7 day moving average:
- (n SGTF) / (n positives) ~ 0.015
- (n B117) / (n SGTF sequenced) ~ 0.95

estimate: 1.4% of positive cases in CA are B117. Trending up.

@EricTopol