Thread by @DarwinAwdWinner

Excited to present our newly published paper, "Molecular states during acute COVID-19 reveal distinct etiologies of long-term sequelae."

Here we identify multiple distinct processes leading to #LongCOVID that have already started during acute #COVID_19.

Tweetorial follows 1/13

It is essential to understand how #LongCOVID (post-acute sequelae of SARS-CoV-2 infection, #PASC) relates to the body's response to #SARSCoV2 infection. We investigated this connection using acute molecular measures and post-acute questionnaires.

rdcu.be/c1f5x

2/13

Important to note, the molecular features (gene expression and anti-spike antibody levels) were measured during acute COVID-19 hospitalization. The same patients filled symptom questionnaires >6 months after discharge. So we're linking two things separated by months/years.

3/13

We looked for gene expression patterns associated with each PASC symptom and whether these patterns matched up with anti-spike antibody levels. We can visualize this with what we call "delta MA plots".

4/13

We looked at gene expression patterns specific to individual cell types (read the methods!). This reveals lots of cell-type-specific expression patterns, already hinting at multiple processes leading to different symptoms.

5/13

Plasma cells are especially interesting. Not only do we see associations with 6 symptoms, but they separate into 2 clusters with opposite expression patterns, mainly driven by antibody-related gene expression, showing 2 different etiologies for these clusters.

6/13

But something is odd here. Remember those antibody levels? The Ab gene expression for many symptoms is independent of the measured Ab levels! How does that make sense? Hypothesis: maybe we're actually observing changes in total Ab levels, not SARS-CoV-2 specific Ab.

7/13

Using another published data set that includes total Ab levels, we found that PASC was associated with differences in total Ab levels but NOT anti-spike (S1) Ab levels. This externally validates that total Ab levels can change without changes in anti-spike Ab.

8/13

In conclusion, we find multiple distinct acute-phase gene expression signatures of long COVID symptoms in blood of hospitalized COVID-19 patients. This:

1. Directly links long COVID symptoms to the body's response to SARS-CoV-2 infection (via anti-spike Ab levels)
...

9/13

2. Demonstrates that different groups of symptoms have different etiologies
3. Raises the possibility of predicting, preventing, and treating long COVID symptoms during acute infection
4. Emphasizes the need to consider the acute phase as we continue studying long COVID

10/13

Thanks to all the patients who provided blood samples and answered the questionnaires, and thanks to all the researchers and staff who contributed to building and analyzing this enormous data set!

11/13

Along with the paper, we have released the corresponding data, which you can access here: synapse.org/#!Synapse:syn3…

Includes:

1. Gene expression (RNA-seq)
2. Anti-spike antibody titers (serology)
3. Extensive clinical data
4. PASC questionnaires

12/13

Our lab is recruiting postdocs! If you're interested in analyzing big complex data sets like this one to answer big complex questions, come work with us! DM me or @noambeckmann1 if you're interested!

13/13