Matt Soellner

Associate Professor, Department of Chemistry @ U. Michigan.

[Matt Soellner @mbsoellner]

@mgdurrant Claude Science and this Claude Team plan is great. But is there a way to couple this (which I have done) to the LSVP? I cannot use Opus 5.5 for kinase inhibitor (chemistry) work targeted at cancer - I applied for LSVP but have not heard back.

[Matt Soellner @mbsoellner]

@Peter_Aldiss @eric_kabrams Agreed. I filled out the LSVP when it first came out and heard nothing. Opus 5.5 is unwilling to do any kinase / cancer work of any sort...

[Matt Soellner @mbsoellner]

@KRHornberger Thanks for posting! The client protein (here c-Src) conformation aspect is intriguing (to me, at least)... Not in this paper, but the DFG-out analog of these PROTACs doesn't degrade c-Src at all (despite being a very good cellular binder/inhibitor). More on that coming soon!

[Matt Soellner @mbsoellner]

@Glaconde34 Thanks for "tweeting" our publication. This was our first foray into PROTACs (quite late to the game), but we have more kinase PROTACs on the way!

[Matt Soellner @mbsoellner]

@london_lab Nanosyn was good (and reasonably priced); bad news. Reaction Biology is great for in vitro. I like Luceome for profiling in lysate and... And Promega (nanoBRET) for cell-based profiling w/ full length constructs.

[Matt Soellner @mbsoellner]

@marwinsegler @D_B_McConnell @felix_s_k My answer (as influenced by Brian Coppola) would be that resonance contributor has one aromatic ring and one ring that is not aromatic. But one can draw a resonance contributor where both rings are aromatic. That is what we teach at UMich…

[Matt Soellner @mbsoellner]

@KRHornberger Good point. At 50-100 µM you will be well past most small molecule CMCs. However, many off-target assays can reasonably be run at 10 µM to assess the polypharmacology. And I'm mostly thinking about this from a 'use inhibitor for biology' approach vs validating a lead compound.

[Matt Soellner @mbsoellner]

@debscience @Chemjobber @AnneJMcNeil @AnneJMcNeil and I use an Amazon Air Quality monitor (no CO2) in addition to the CO2 meter above. Cooking (whether gas or stove/griddle) is always apparent with a drop in AQ (due to spike in PM and VOCs). Yesterday's AQ (you can see when we made meals):

[Matt Soellner @mbsoellner]

@HartungIngo Here's my totally biased suggestion: https://t.co/Y06zEKqvyn

[Matt Soellner @mbsoellner]

@JSheltzer PLCG2 mutation (gain of function) is a common resistance pathway for BTK inhibitors. But often alongside BTK mutation (which might exclude this example from your criteria).

[Matt Soellner @mbsoellner]

@drugmonkeyblog One more. The R00 at least used to be a total cost grant. I had no idea the implications of that until I had destroyed my year 1 budget. Then I ended up trying to buy as much equipment on it as possible. Because equipment = no overhead. Hopefully they have fixed that with R00s.

[Matt Soellner @mbsoellner]

@drugmonkeyblog That every R01 isn’t at least $250K a year direct. Some ICs can / do cut up to 24% of the budget (even when modular budget) without allowing changes to the aims. (25% cut and you can renegotiate). That was a shock to me. Until it happened on a modular R01 budget.

[Matt Soellner @mbsoellner]

@JSheltzer @CellDeathLab And per DepMap, SAOS2, SKBR3, and SUM159 are not very EGFR dependent. Assuming knockdown correlates with inhibition (not always true true with kinases) I would expect that a selective p38 inhibitor would work in these cell lines via p38 inhibition and not via EGFR inhibition.

[Matt Soellner @mbsoellner]

@JSheltzer @CellDeathLab Within DepMap there is pretty good correlation of MAPK14 dependence between their CRISPR and RNAi datasets. A few common cell lines with MAPK14 dependence in both data sets.

[Matt Soellner @mbsoellner]

@JSheltzer Of course same is true for kinase inhibitors with multiple targets. Many examples of context-dependent activity where 2D vs 3D vs choice of media has a huge impact. With a “selective” p38 inhibitor in a context without p38 activity dependence, off targets will drive efficacy.

[Matt Soellner @mbsoellner]

@JSheltzer MAPK14 dependence is likely context dependent. DepMap data lists MAPK14 dependencies. We’ve seen MAPK14 dependence when specific cell lines (HT-29) are grown in 3D culture using physiological media.

[Matt Soellner @mbsoellner]

@JSheltzer At the cellular IC-80 of kinase Y, what is the percent inhibition (cellular) of kinase Z?

Assuming Hill slope = 1.0, IC-80 for kinase Y = 20 nM. Thus, kinase Z would be ~30% bound/inhibited.

While semantic, I would not call that "highly selective".

[Matt Soellner @mbsoellner]

@aledmedwards You are assuming that the "public information" (which accounts for 1/3 of leads) does not include HTS campaigns?

Regardless, academic HTS can be a teaching tool for trainees. It can also be useful to demonstrate "druggability" for new targets.