Question for epidemiologists: in your field, is it acceptable to reverse engineer data out of a figure (left) using Illustrator to generate a nicer version (right) for your paper?

Sources:
who.int/docs/default-s…

zenodo.org/record/5075888…
The original data cannot be accessed, legend is blurred out, the map itself is highly pixelated.

But would it be better to copy the original figure into one's new paper (with attribution) alongside the reverse-engineered figure so that readers are aware of the 'data' quality?
The authors do clarify in their supplementary (but not main text) that data loss occurred during reverse engineering:

"Map data was manually extracted from Fig 17.. using Adobe Illustrator. Because of multiple overlapping points there will be errors in the extraction process."
Because I think if I ever tried to reverse engineer data points out of a blurry volcano plot to generate a figure for my paper, the peer reviewers would have a field day.

• • •

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

Keep Current with Alina Chan

Alina Chan 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 @Ayjchan

9 Aug
On the question of whether covid-19 might’ve come from Europe.

TLDR it didn’t.
Also this is really how peer review should work. Open, honest and fair criticism of preprinted/published work. Now everyone can see what problems each study suffers from and decide for themselves whether they think the conclusions are well supported. Expertise is being shared.
On studies pointing to covid appearing in Europe earlier than Dec 2019, many somehow detect later variants (D614G) rather than the earliest form of SARS2.

This requires a virus endemic to South China/SE Asia to not only be able to time travel but also teleport across continents.
Read 9 tweets
8 Aug
If anyone needs a good 😂 this is the figure from the China-WHO report showing the home address of covid-19 cases from Dec 11, 2019 or earlier.

A team of scientists, unable to access early case data, had no choice but to extract information from this fig using Adobe Illustrator.
Problematically, despite being already of such low resolution, the map still has 'unintended errors' as revealed by @washingtonpost
washingtonpost.com/world/asia_pac…
These errors are shown here in a diagram made by @InWuchang
Read 17 tweets
7 Aug
I'm seeing a lot of confusion about this recent @CNN story about "reams of genetic data from Wuhan lab".

Nothing in the article specifies that this is the WIV pathogen database. It could just as likely be raw data deposited into public databases.

cnn.com/2021/08/05/pol…
(1) "genetic blueprints drawn from virus samples studied at the lab in Wuhan"

No time range specified. Could've even been post-covid.

No lab specified either. Could be just Wuhan labs in general.
(2) "the machines involved in creating and processing this kind of genetic data from viruses are typically connected to external cloud-based servers"

With the WIV pathogen database, it is unclear how many samples were sent for NGS, but my guess would be not a large %.
Read 7 tweets
6 Aug
I really like the idea of re-routing viral traffic, put forth by virologist Stephen Morse.

Natural routes of viral traffic continue to exist and be amplified. But new routes of viral traffic, related to research activity, have emerged in the modern era...
noemamag.com/the-routes-of-…
One early example of a novel pathogen emerging due to research activity is Marburg virus, "first recognized in 1967, when outbreaks of hemorrhagic fever occurred simultaneously in laboratories in Marburg and Frankfurt, Germany and in Belgrade, Yugoslavia"
cdc.gov/vhf/marburg/in…
And that was generations ago.

Today, we can go from a virus sample to a genome sequence in a few days, and from a genome sequence to a completely synthetic perfect (seamless) copy of that genome in a matter of weeks.
Read 5 tweets
31 Jul
Would be good to track the 2000+ individuals (all adults vaccinated) from this cruise who are disembarking in the Bahamas tomorrow and see how many test positive for Covid-19 over the next 2 weeks. usatoday.com/story/travel/c…
There are many things we don’t know *yet* about the delta variants and vaccines:

How much do our vaccines protect against infection, asymptomatic or symptomatic, by delta?

How much do infected vaxed individuals spread the virus compared to infected non-vaxed individuals?
We do know that several top vaccines continue to protect against severe covid-19. So it’s not particularly worrying when vaxed people get delta (although there are exceptions and some breakthrough cases suffer greatly). What’s worrying is the spread of delta to the unvaxed.
Read 7 tweets
30 Jul
NIH explained to Senator Grassley how NIH-funded SARS/MERS-like chimeric CoV work at WIV had not been determined to meet USG criteria for GOFROC covered by federal funding pause or P3CO Framework.

But more interesting than the ⚔️ on GOF definitions is…
grassley.senate.gov/imo/media/doc/…
“during the course of the grant, the grantee proposed to place a small portion of the newly identified bat coronaviruses into a larger portion of MERS-CoV to understand the potential origins of MERS-CoV in bats.. conducted at WIV”

Can this proposal pls be shared with the public?
Another question: Did NIH-funded WIV work turn out to be “instrumental to the unprecedented rapid development of vaccines, therapeutics, and diagnostics to address the COVID-19 pandemic”?

Important for this to be substantiated with data since similar work continues to be funded.
Read 7 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

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

Donate via Paypal Become our Patreon

Thank you for your support!

Follow Us on Twitter!

:(