Kava Science Kava as GABAergic? Not so fast.
- Thread starter The Kap'n
- Start date
Forretressqt
Kava Enthusiast
The emperor returns!
I know I sorta just slid in there like nothing happened, lol. Yeah, I'm getting my gears back in place and getting back online.
kasa_balavu
Yaqona Dina
Bula JP
KavaTasteGood
Kava Curious
Hi Kaptain,
A few years ago I posted here about trying to locate the binding site of kavalactones using (relatively fast) computational methods. I ended that post by noting that a far more expensive method could give a more definitive answer. I've finally applied that method — unbiased coarse-grained molecular dynamics — to find the binding site of kavain on the GABA-A receptor.
A couple of things before the result. The analysis isn't finished yet, and two caveats matter here: (1) the method is new to me — I'm not at expert level and may have made some parametrization errors; and (2) the method itself has limitations. It's essentially a trade-off, giving up a bit of resolution for a big speed-up in the calculations — something you need for a system as large as the GABA-A channel embedded in its membrane, especially with the protocol I ran (unbiased MD to sample a kavain binding event). To give you a sense of the cost: even with this enhanced approach, I simulated for several weeks on two GPUs that cost ~€10,000 each.
With those caveats in mind: my extensive simulations don't converge on a clear binding site for kavain on the GABA-A receptor. To place this — assuming I didn't fuck up somewhere — in silico methods like this one fall between 2 and 3 on your evidence-strength table. I'd call it 3a: definitely not mere rhetorical reasoning, but weaker than in vitro evidence. Regardless of strength, though, in silico methods can be great for illustrating and understanding in vitro results.
I was expecting binding at the propofol site, where the N265M point mutation affects both propofol and kavain activity (see Chua et al., 2016). The plan was to run the protocol on both the wild-type receptor (WT, i.e. non-mutated) and the N265M mutant and compare the two — but binding never occurred in the WT.
A few years ago I posted here about trying to locate the binding site of kavalactones using (relatively fast) computational methods. I ended that post by noting that a far more expensive method could give a more definitive answer. I've finally applied that method — unbiased coarse-grained molecular dynamics — to find the binding site of kavain on the GABA-A receptor.
A couple of things before the result. The analysis isn't finished yet, and two caveats matter here: (1) the method is new to me — I'm not at expert level and may have made some parametrization errors; and (2) the method itself has limitations. It's essentially a trade-off, giving up a bit of resolution for a big speed-up in the calculations — something you need for a system as large as the GABA-A channel embedded in its membrane, especially with the protocol I ran (unbiased MD to sample a kavain binding event). To give you a sense of the cost: even with this enhanced approach, I simulated for several weeks on two GPUs that cost ~€10,000 each.
With those caveats in mind: my extensive simulations don't converge on a clear binding site for kavain on the GABA-A receptor. To place this — assuming I didn't fuck up somewhere — in silico methods like this one fall between 2 and 3 on your evidence-strength table. I'd call it 3a: definitely not mere rhetorical reasoning, but weaker than in vitro evidence. Regardless of strength, though, in silico methods can be great for illustrating and understanding in vitro results.
I was expecting binding at the propofol site, where the N265M point mutation affects both propofol and kavain activity (see Chua et al., 2016). The plan was to run the protocol on both the wild-type receptor (WT, i.e. non-mutated) and the N265M mutant and compare the two — but binding never occurred in the WT.
Last edited: