hERG binding correlates with LogP?

Here is a very good analysis from human ERG blockers article:

The plot below (created usingVortex)shows pIC50 calculated from the literature IC50 data versus calculated logP determined usingalogPS, the colour coding shows the overall general tend of increasing hERG activity as logP increases, it also highlights the reduced liability seen with acids (green) and zwitterions (red).



Whilst the majority of data is derived from radioligand binding experiments (using either Dofetilide or MK-499), there is a substantial amount of data from patch clamp experiments, I collated enough data now (covering 4 orders of magnitude) to give an idea of how the assays compare. As you can see there is a reasonable correlation between the assays, but there are one or two outliers. Which is more predictive of in vivo activity is an excellent question that I don’t have the data to answer yet.



I still need to increase the size of the data-set, and if anyone can direct me to any publicly available data, or to publications that contain data i'd much appreciate it.
Worth reading, Medicinal Chemistry of hERG Optimizations: Highlights and Hang-Ups, Jamieson, Journal of Medicinal Chemistry, 2006, 5029

q-hERG: QUANTUM's innovative approach to hERG binding calculations is updated to v 2.0

Quantum Pharmaceuticals, the owner of this blog, is proud to release version 2.0 of its innovativ HERG protein binding prediction software.



QUANTUM hERG (q-hERG) screening assays is a unique and innovative computational approach, which allows you to predict from a molecule structures of compounds their inhibition constants (IC50) for hERG channels.

q-hERG: QUANTUM's innovative approach to hERG binding calculations is finally released

QUANTUM hERG (q-hERG) screening assays is a unique and innovative computational approach, which allows you to predict from a molecule structures of compounds their inhibition constants (IC50) for hERG channels.

q-hEARG features:

• Output is pIC50 values (-logIC50) for the molecules. The accuracy of prediction is 1.1 pIC50 units;
• No training sets or QSAR methods applied;
• hERG inhibition prediction is made by docking of compound on Quantum Pharmaceuticals’ Proprietary Flexible 3D structure of hERG;
• Docking is based on quantum and molecular physics (see Quantum Science Core for an overview);
• Average correlation has RMSD=1.18 pIC50 unit, and correlation coefficient = 0.82;
• Easy to use user interface, no special hardware requirements, both Linux/Windows supported;
• You can also request services based on QUANTUM hERG Screening Assays.

q-hERG is an independent software module, sharing the user interface and basic usage concepts with our q-ADME: ADME/PK properties prediction software, q-Mol: physico-chemical properties calculator, and q-Tox: toxicological profiling software. More information, including q-hERG product booklet can be obtained from the Quantum Pharmaceuticals products site.

Obtaining Q-Albumin software:

Please review your licensing options, add Q-Albumin: QUANTUM Albumin Binding Prediction Software to your shopping card and checkout to get the download links.

Licensing Options: One Month License $50,00 One Year License $100,00

And continue to CHECK OUT

HERG binding prediction quality: q-HERG model vs. experiments

Quantum Pharmaceuticals has recently completed development of its in-house HERG-protein binding model. Since there is no 3D structure of HERG-protein available, the calculations envolved a number of fits and model assumptions.

To see whether our data is notoverfitted, we compared the errors inour calculations with experimentaluncertanty of binding affinities for the same set of molecules. The graph on the left shows two sets of points: q-HERG model vs. experiment (red squares) and pIC50 values for the same molecules taken from different sources (green triangles, see our How good are biological experiments? HERG binding data analysis post for more details).

The two distributions are roughly of the same width, which, in a way, provides a sanity check for our HERG model.

How good are biological data - II: Trombine, GSK, GPCR

Many bindign affinity prediction methods, such as scores and QSAR models, rely on availability of accurate information on binding constants. The figure on the left is a result of our sdf-file parser applied to trombine (blue) and GSK (yellow) binding data from BindingDB database. The parser is written with python and uses pybel to extract unique molecules from a given multimolecular sdf.
The parser not only finds identical (in Tanimoto-similarity sense) compounds, but also prints the binding constants from the sdf records. The graph shows the correlation of the reported inverse log(binding constants) for the same molecules from different entries (sources).
The result is in fact fairly impressive (the blue points): the discrepancies-"errors" are quite large and are especially profound for good (or better say very good) binders.
The yellow points represent the result of the same script over GSK-kinase activity data. Although the total number of molecules in BindDB is much larger, almost all of them are unique. The difference between different sources is not as much as for trombine.
The Figure on the right is the visualized script output for GPCR(5-HT2B) from PDSP Ki database. The situation is roughly the same: the accuracy of a typical biological experiment reported in a literature amounts roughly to a single unit of pKd.

This and previously reported correlation for HERG ion channel should serve as an example when the results of binding affinity calculations are compared to experimental data.

How good are biological experiments? HERG binding data analysis

A correlation between predicted and expermentally measured values of biological activity is a natural measure of a model quality. For instance, QUANTUM docking software calculates binding free energies, which are directly comparable with experimental values of -p(binding constant, Kd). Root mean squared error between the measured and the calculated quantities is the quantitative measure of the software performance.
Whatever the correlation is presented to prove the validity of a model, another important issue is the quality of the experimental data itself. The reported values for binding constants (or activities) often vary because of different measurement strategies, experimental errors or interpretation uncertanties. To visualize the situation we investigated a few datasets for HERG binding taken from QSAR World website.
The downloaded files were saved in source folder and processed with the following simple python script (thanks to openbabel):

files = os.listdir('source/')
molecules = []
for file in files:
molfile = readfile("sdf",'source/'+file)
for mol in molfile:
molfp = mol.calcfp()
present = 0
for savedmol in molecules:
savedmolfp = savedmol.calcfp()
if (molfp | savedmolfp == 1):
present = 1
print mol.data, savedmol.data
if (not present):
molecules.append(mol)

The results where analyzed in a spreadsheet program and represented on the graph above. A lot of molecules occur multiple times in the datasets. While in many of the cases the activities coinside up to 0.01 (which most probably indicates citing from a single source), the remaining values thouch correlated with each other, differ by roughly a single pKd unit.