How different energy between unbound and bound conformation of molecules? #memo #journal #JCIM

As we know that, ligand conformation is very important factor of drug design and also it’s difficult to predict binding pose of ligand. So ligand conformational strain energy(LCSE) is very important factor.

In the LBDD, we can’t know ligand binding pose so, CompChemists try to generate 3D conformation with several ligand based techniques such as MM, QM, etc. But how different structure between QM based global minima and bound state? Recently I found very interesting and useful articles in JCIM. Here is an ULR.
https://pubs.acs.org/doi/10.1021/acs.jcim.0c01197

The authors surveyed LCSE of large amount of protein-ligand data set with high level QM calculations such as M062X-D3/ma-TZVPP (SMD)//M062X-D3/def2-SVP(SMD).

At first they curated data set, for example not covalent binder, common elements in drugs etc… Details are described in Fig1.

After the data preparation, they conducted QM calculations against the data and disclosed that the mean of strain energy was 4.6 kcal/mol and median was 3.7 kcal/mol. I surprised that the mean value of LCSE is not so small it means that binding conformation is strained. Strain Energy is defined below.
Estrain = Ebound – Eglobalminimum

After the figure, the author analysed data more deeply. And Fig4(e) shows an example of large change of bound and unbound conformation of ligand PDB code is 4Y2H. The ligand of 4Y2H takes extended form in bound form but not extended form in unbound form due to intramolecular interaction.

It’s worth to know for me only ligand based global minimum structure isn’t same as binding form because it can’t consider ligand-protein interactions such as cation-pi, pi-pi, etc…

So LBDD with 3D structure is still challenging field I think.

In the publication there are more discussions about LCSE. If reader who has interest it I recommend to read the article.

Advertisement