Thread by @guilbourque

1/ We've just finished 3 papers on #transposons #infection #epigenomics #graphGenomes, the result of an amazing collaboration with @LB_Barreiro and Tomi Pastinen. Start of a long 🧵

2/ The first paper by lead author @katiearacena is:

Epigenetic variation impacts ancestry-associated differences in the transcriptional response to influenza infection
tinyurl.com/ytvrk5hn

3/ @LB_Barreiro just talked about this paper at #BoG22 and @katiearacena has a poster so I'll let them explain their results!

4/ The second paper by lead author @xunchen85 is:

Transposable elements are associated with the variable response to influenza infection
tinyurl.com/4mp6exhu

5/ The third paper by lead author @groza_cristian is:

Genome graphs detect human polymorphisms in active epigenomic state during influenza infection
tinyurl.com/yc4aj6mp

6/ There are so many cool results to describe in these papers that I'm not sure where to start! I'll be back with some of the key highlights soon.

7/ These studies are part of @CEEHRC @IHEC_Epigenomes and were possible thanks to funding and support from @CIHR_IRSC @GenomeCanada @CalculQ @Alliance_Can @McGillGenome

8/ The paper on #transposons and #immunity with @LB_Barreiro started with an interesting observation. Although we infected macrophages from 39 individuals with #influenza in the same way (study described at #BoG22), the number of viral reads varied tremendously

9/ We could show that this was also associated with differential activation of immune pathways. Now the question was, are #transposons contributing to this variability?

10/ First, we did the classic analysis of looking for #transposons families that were enriched in active chromatin following infection. But, because we had many individuals, we had ranges of enrichment and take a look at families like LTR12C

11/ Some families were highly variable in their enrichment while others were more stably enriched. Could this help explain the variable response?

12/ Small aside here but we could also look at the reverse, families with decrease accessibility following infection. Also very interesting, with enrichment mostly in #L1 but no variability here

13/ Back to the variable enhanced families, here's what it looks like if we compare a low var family (THE1B) with a high var family (LTR12C). Each line here is a different individual and we see how much it varies

14/ A different look at this shows that there are really two types of #transposons families that become accessible following infection. What we call #highvar and #lowvar families

15/ The paper includes lots of other analyses showing how these regions also have other epigenetic marks, are close to regulated genes, etc. But i'll skip to the motif analysis that really helped us understand the difference between #highvar and #lowvar.

16/ To set the stage, look at this great figure made by @xunchen85. We have so much data (35 individuals) that we can really pin point the peaks in each instance of each #transposons family. For many families, there's evidence of multiple regions opening up.

17/ But the really exciting result was that using these precise peaks we could identify completely different motifs in #highvar versus #lowvar. #lowvar had many of the well known immune regulators but #highvar had very different proteins including many #KRAB_ZNFs

18/ I'm skipping some of the details again but in the end, we could show that using #transposons, and factors regulating #transposons, we could predict viral load (remember tweet #7 😉) from the baseline very well

19/ That's my "quick" summary of the second paper. We see major shifts in #transposons following #influenza infection and, interestingly, we see a lot of inter-individual variability which appears to be associated with the variable response (at least as measured by viral reads).

20/ I really want to congratulate my postdoc @xunchen85 who worked tirelessly on this in my lab in @Ashbi_KyotoU since the beginning of the pandemic. I can't wait to come to #Japan at the end of the month so that we can celebrate in person!

biorxiv.org/content/10.110…