meyer-lab/gc-valent

Plots for pre-clinical paper with Visterra

Closed this issue · 2 comments

aarmey commented

Just wanted to record a few notes about my discussion with Scott. We settled on a couple plots that would help integrate the model with what they want to show.

Figure 1: A model of specificity using the Treg:Thelper and Treg:NK ratios.
This is like a plot we've prepared for them before. It should be as simple as possible and so we figured the ratio of pSTAT5 was better than the specificity metric in the paper.

Supplement: Predicted vs. actual for monomeric H16N and H16L.
We don't need to show naive vs. memory. This can be structured like the predicted vs. actual comparisons in Farhat et al. (in the supplement). We'll also need to identify a way to show this in condensed form within the main figures. The condensed version could just be a couple populations, reduced set of timepoints, or something else...

borcuttjahns commented

Sounds good! A couple quick questions:

  1. Is this all assuming we use the updated or old ODE model? Or would we think about using the GP model?
  2. Regarding Figure 1) - when you say "a model of specificity" and mention that the ratio of pSTAT5 should be used I'm a bit confused. Would this be experimental or predicted measurements of pSTAT5 responses over a dose response curve? And isn't the ratio of pSTAT5 in Tregs vs. Thelpers and NK cells what we used previously in the paper as our specificity metric? Or is there a difference in what you want to use now?
  3. Is there a timeline we need to follow for getting this back to Scott?

Thank you very much!

aarmey commented
  1. Ideally this is using the old model. The ODE model is mechanistic and so can be expected to more accurately extrapolate from those conditions we have observed. We also figure this makes writing this new paper easier—we can just reference Farhat et al.
  2. Remember that we had made some plots with varied IL-2Ra and beta/gamma affinity, with the Treg/Thelper and Treg/NK pSTAT5 ratio as the y-axes? It's not from Farhat et al but from some internal Visterra presentations. We can discuss it more tomorrow if helpful.
  3. This isn't urgent. I told Scott we'd plan on sending something in ~3-4 weeks.