Tuesday, December 30, 2008
How much water is in a Generalized Born protein?
In practice applications of Generalized Born models are further complicated by various approximations for calculating volume (or surface) integrals, removing atom overlaps etc. What remains left is some sort of approximation to molecular volume (surface) and the so called Born Radii for every atom.
Each of the Born radii quantitatively shows a degree to which an atom is "buried" within the protein. The presented graph gives a simple idea to a which extent GB can even be used for description of solvation energies of a simple, model spherical protein containing approx. 1000 atoms of carbon.
The red squares give the dependence of the Born Radii on the atom positions. The points are obtained using our own implementation of AGBNP, one of the best realizations of GB procedures available in the literature.
The yellow curve represents exact result for a spherical protein, where GB and exact analytical expressions coinside. As one can see, AGBNP result fails to grow inwards and saturates at a very small value at r=0.
The reason for this behaviour is two-fold: first AGBNP is based on the so-called Coulomb approximation and thus can not be exact. Indeed, Coulomb approximation fails at the protein boundary and gives d(Born Radius)/dr twice as large as the exact result. This is a true problem, but it can not explain fundamentally wrong results in the protein center!
The other problem of AGBNP (and in fact any GB model), is that the model implies a certain approximation for molecular surface and the surface may have water filled cavities inside the protein! The cavities represent (within the same model) a medium with high dielectric constant and decrease the value of the Born radii.
To check the last assumption we searched for the water filled cavities removed them (to a certain adjustable extent). The result is represented by the blue circles and shows a clear improvement towards reproducing the exact analytical result.
Conclusion? Dry your protein up before even attempting to use GB approximation to get a good solvation energy for a large molecule!
Tuesday, December 16, 2008
Tuesday, November 25, 2008
2008 Quantum's technology platform update and software releases
The progress in our understanding of biomolecules interactions led to further accuracy improvements in our major calculation routines (IC50, solvation energy, etc.). Speed increase and sophistication of the models employed in our simulations provided better ways for false positive elimination. Direct application of our software brought up novel inhibitors of HIV integrase and gp120 proteins, human neutrophyle elastase (HNE) (see collaborations). Massive computations made using Amazon EC2 computing platform let us develop new and refind existing ADMET models (see drug absorbtion prediction (arXiv:0810.2617) as an example).
All the scientific advances are plugged in and available through the following releases of Quantum sofware (sold separately and in packages at discount prices):
q-TOX - enables researches to compute toxic effects of chemicals solely from their molecular structure (LD50, MRDD, side effects) . The robust model is based on completely new approaches. While there are numerous commercially available toxicity prediction software, none offers the depth, scope and precision comparing to q-TOX. The paradigm in the q-TOX approach is based on the premise that biological activity results from the capacity of small molecules to modulate the activity of the proteome.
q-Mol - calculates such physicochemical parameters as Solubility in H2O (g/l); Solubility in DMSO (g/l); LogP, water/octanol; Mol weight; H-bond donors; H-bond acceptors; The number of rotatable bonds;Lipinski-rule-of-5.
q-ADME - For the first time we identified proteins, binding to which correlates well with FA and T1/2. This enabled us to simulate the active component of the ADME properties that has been the heel of Achilles for existing computational approaches still. The software predicts the following properties: Drug half-life (T1/2); Fraction of oral dose absorbed (FA); Caco-2 permeability; Volume of distribution (VD); Octanol/water distribution coefficient (LogP)
q-hERG - a unique and innovative software, which allows you to predict from a molecule structures of compounds their inhibition constants (IC50) for hERG channels.
q-Albumin software takes a molecular structure and calculates HSA binding constant by docking the molecule to both of the HSA active sites (Sudlow site I and Sudlow site II).
Monday, November 3, 2008
5th of November talk@MIPT Interdisciplinary Seminar "Water as a ferroelectric: anomalous properties, long range order and interactions of nano-par....
Moscow Instutite of Physics and Tehcnology, November 5th, 2008. "Water as a ferroelectric: anomalous properties, long range order and interactions of nano-particles in solution" (in russian)
The presentation will be held in room 202НК, 18:35 (read full announcement here).
Tuesday, October 28, 2008
Quantum Pharmaceuticals enters collaboration with Children's Cancer Institute Australia
Moscow, October, 28 2008
Quantum Pharmaceuticals announce drug discovery collaboration with Children's Cancer Institute Australia's (CCIA). Under the terms of the agreement Quantum Pharmaceuticals gets access to CCIA in-house disease target data. Quantum Pharmaceuticals will contribute its technological breakthroughs and expertise in small molecule drug discovery to feed the portfolio of CCIA with new drug candidates. CCIA is to further develop the discovered inhibitors. The targets and financial terms were not disclosed.
About Quantum Pharmaceuticals
Quantum Pharmaceuticals is a drug discovery company based in Moscow, Russia specializing in small molecule screening and design through the use of its proprietary technology platform.
About CCIA
Children's Cancer Institute Australia's (CCIA) vision is to save the lives of all children with cancer and eliminate their suffering.Our mission is to be a leader in preventing cancer, to find new ways of curing cancer in children through world-class research, to ensure the best possible quality of life for these children and their families, to share the vision with others and to increase awareness, participation and funding.
Quantum Pharmaceuticals collaborates with University of Pittsburgh on HIV drug discovery.
Moscow, October, 20 2008
Quantum Pharmaceuticals and University of Pittsburgh announced a drug discovery collaboration in HIV sphere. Under the terms of agreement Quantum Pharmaceuticals gets access to the target data from University of Pittsburgh. Quantum Pharmaceuticals will apply its industry leading computational technology to discover novel small molecule inhibitors for this target. The University is to provide biological expertise and further develop the discovered inhibitors. The financial terms of the deal were not disclosed.
About Quantum Pharmaceuticals
Quantum Pharmaceuticals is a drug discovery company based in Moscow, Russia specializing in small molecule screening and design through the use of its proprietary technology platform.
About University of Pittsburgh
Founded in 1787 the University of Pittsburgh has evolved into an internationally recognized center of learning and research. The University’s 12,000 employees, including 3,800 full-time faculty members, serve about 34,000 students through the programs of 15 undergraduate, graduate, and professional schools.
Friday, October 10, 2008
q-hERG: QUANTUM's innovative approach to hERG binding calculations is updated to v 2.0
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.
Wednesday, October 8, 2008
Making a good water model: Molecules do conformationally change when cross from a gas to water solution
The graph on the left represents two types of solvation energy calculations compared with experiments. The first series (small circles) are the energy differences on solvation for a set of molecules without conformational changes taken into account. The second set (large squares) is obtained after a single optimization run.
The correlation with the experiment clearly improves after conformational changes calculations. Apparently this does not only mean that the model is good, it also means that the molecules do change structure when inserted into water from the gas phase.
Thursday, September 25, 2008
From Biological Spectra (multiple protein binding data) to pharmacological profiling!
Even if the efficacy is good, another important property of a compound is its toxicity. Toxicity is related to the compound physical properties, such as solubility etc, as well by its ability to bind to and hence inhibit various vital human proteins (and may be even DNA and RNA).
Common sense suggests that an ideal compound binds its specific drug related target and does not bind to anything else. Anything in between is toxic, at least to a some extent. For example, most of important properties utilize ATP molecules, which means that human body contains a lot of ATP-bindig proteins. If you make a drug attacking an ATP-binding site of a "bad" protein, most probably, a lot of "good" and useful proteins will be also affected. In that case your compound should be toxic. This is indeed the case for many cancer drugs attacking ATP-binding sites of kinases.
The latter statement is the foundation of our approach. Although it's quite conceptually simple, it's useless unless it can be supplemented by a meaningful mathematical model. Let us dwell into some more details to see how the whole thing can be made working.
Let us overview important properties of a drug candidate. First there is a bunch of physical properties, such as solubility, differential solubility, LogP (namely the difference between water and lipid solubility) etc. These quantities are easy to measure, are of direct physical meaning and can be pretty easily calculated (with or without QUANTUM software).
Another set of characteristics defines a compound ability to penetrate through cell membranes and its biochemical in liver. These are quantities deturmining bioavailability, half life, volume of distribution etc. None of such quantities can be evaluated using the simple physical properties alone. For example, drug absorbtion depends on the molecule interaction with proteins actively transporting the molecules through the cell membranes.
The bottom line: bioavailability and other quantities require understanding of a compound binding properties to a selected number of proteins participating in a compound transport and metabolism.
So the conclusion is that IF YOU KNOW WHICH PROTEINS ARE IMPORTANT, AND IF YOU CAN CALCULATE HOW YOUR COMPOUND BINDS TO THEM, YOU KNOW THE COMPOUND PHARMACOLOGICAL AND TOXICOLOGICAL PROPERTIES
Now the only problem how to identify those "important" proteins.
Fortunately, there are thousands of molecules with known properties. What we can do is the following:
- take a molecule
- calculate its binding to any human protein with known 3d structure
- use the obtained binding affinities (numbers) as a molecule's binding profile fingerprint (the Biological Spectrum), characterizing the ability of the molecule to interact with the whole human proteome
Now assume we know such Biological Spectra for 1000s molecules with well known properties. This means we can now datamine the fingerprints->known properites relations. The basic premise is, of course, that the molecules with similar fingerprints have similar properties.
We have a number of proofs of such technology working. The most recent one is the prediction of active transport drug absorption properties for drug like molecues based on the binding data against human brain hexokinase type I-related protein. We prove that the binding energy of a compound against the protein may serve to distinguish between the passively and actively transported molecules and even help to calculated the drug absorbtion quantitatevely.
From binding data to pharmacokinetics: a novel approach to active drug absorbtion prediction
To this end, several kinetic models for drug absorption within the body have been introduced (see e.g. ). They necessarily introduce some simplifications belonging to the category of the so-called three-compartment models where the substances (such as drugs or nutrients) move between three volumes (e.g. the human organs). In fact the models require two kinds of molecular properties. First are purely physical characteristics, such as solubility, differential solubility, LogP etc. These quantities are easy to measure or to calcualte, have direct physical meaning and sufficient to predict absorbtion profile of passively absorbed drugs. Actively transported molecules interact with protein transporters and therefore prediction for actively transporting compounds require a lot of separate knowledge of binding to and kinetics of the transporting proteins.
The major objective of this investigation was to develop a drug absorbtion prediction approach based on entirely different paradigm, thus avoiding difficulties of both knowledge-based and QSAR-based models, and therefore capable of better predictions. Recently it was observed that experimental values of molecular activities against a large proteins set can be used for predicting broad biological effects . In this investigation we take advantage of this concept and develop a novel quntitative method for identification of actively transported drugs. To do that we performed a docking study of a few hundreds small molecules (mostly drugs) against a diversified 510 proteins set representing human proteom. Using available absorbtion data for each of the molecules we obtained a support vector classifier capable to identify proteins which affinity for drugs correlates well with the active absorption of these drugs in 81% cases. The observation helped us improve our passive absorbtion model by adding non-liner fluxes associated with the transporting protein to obtain also a quantitative model of the passively absorbed drugs.
Ref: arXiv:0810.2617 [ps, pdf, other]
- Title: From protein binding to pharmacokinetics: a novel approach to active drug absorption predictionComments: 9 pages, 5 eps figuresSubjects: Quantitative Methods (q-bio.QM); Biomolecules (q-bio.BM)
The nature of percolation phase transition in films of hydration water around immersed bodies.
Reference: accepted for publication to Journal of Structual Chemistry (Russian Journal of), 2008
Spontaneous polarization of a polar liquid next to nano-scale impurities
The spontaneous polarization of liquid induced by the immersed in it nano-size bodies (proteins, peptides, …) results in the additional long-range interaction between them that depends on their relative orientation. Polarization of liquid in this case looks like that presented in Fig.1 in agreement with MD. All mentioned MD results can not be explained in frames of standard continuous scalar theory of water. These phenomena were analyzed here in frames of continuous vector model of polar liquids applications of which looks like promising to speed the simulations of macromolecular complexes.
Reference: arXiv:cond-mat/0601129 [ps, pdf, other]
- Title: Long-Range Order and Interactions of Macroscopic Objects in Polar LiquidsComments: 11 pages, 6 figuresSubjects: Soft Condensed Matter (cond-mat.soft); Chemical Physics (physics.chem-ph); Biomolecules (q-bio.BM)
Accepted for publication in Journal of Physical Chemistry A (Russian Journal of), 2009
What's an ultimate value of reversible drug binding constant?
To demonstrate the existence of the ultimate value of the binding energy two models are considered: macroscopic and microscopic, both giving the same conclusions: the critical value of absolute meaning of binding energy is 50-100kJ/M. If the binding energy exceeds this value, then drug essentially perturbs protein configuration. In a microscopic picture this perturbation is a sequence of irreversible conformational transitions in protein body. In a macroscopic one it is an inelastic deformation of a protein substance. Our estimation agrees with the experimental value (50 kJ /M) of the ultimate energy that can be stored in a protein molecule without its destruction.
The existence of the critical value of binding energy should be accounted in structure based drug design methods where protein molecule is considered in an elastic deformation approximation.
Reference: accepted in Russian Journal of Biophysics, 2008
Wednesday, August 20, 2008
Quantum LogP module (part of q-Mol package) has been benchmarked by vcclab.org
"Quantum LogP, developed by Quantum Pharmaceuticals, uses another quantum-chemical model to calculate the solvation energy. Like in COSMO-RS, the authors do not explicitly consider water molecules but use a continuum solvation model. However, while the COSMO-RS model simplifies solvation to interaction of molecular surfaces, the new vector-field model of polar liquids accounts for short-range (H-bond formation) and long-range dipole–dipole interactions of target and solute molecules Quantum LogP calculated log P for over 900 molecules with an RMSE of 0.7 and R2 of 0.94".
Monday, August 18, 2008
Ferro-electric phase transition in a polar liquid and the nature of lambda-transition in supercooled water
In the our latest publication, Ferro-electric phase transition in a polar liquid and the nature of lambda-transition in supercooled water, we develop two related approaches to calculate free energy of a polar liquid. We show that long range nature of dipole interactions between the molecules leads to para-electric state instability at sufficiently low temperatures and to a second-order phase transition. We establish the transition temperature, T_{c}, both within mean field and ring diagrams approximation and demonstrate that the ferro-electric transition is a sound physical explanation behind the experimentally observed \lambda-transition in supercooled water. Finally we discuss dielectric properties, the role of fluctuations and establish connections with earlier phenomenological models of polar liquids.
Reference: arXiv:0808.0991 [ps, pdf, other]
- Title: Ferro-electric phase transition in a polar liquid and the nature of \lambda-transition in supercooled waterComments: 4 pages, 1 eps figureSubjects: Statistical Mechanics (cond-mat.stat-mech); Soft Condensed Matter (cond-mat.soft)
Sunday, June 15, 2008
Quantum Pharmaceuticals announce collaboration with University of Colorado at Boulder
Moscow, July 15 2008
Quantum Pharmaceuticals announce drug discovery collaboration with University of Colorado at Boulder. Under the terms of agreement Quantum Pharmaceuticals will apply its state-of-the-art in-house drug discovery technology to discover novel small molecule inhibitors in inflammation area. CU-Boulder is to further develop the discovered inhibitors. The targets and financial terms were not disclosed.
About Quantum Pharmaceuticals
Quantum Pharmaceuticals is a drug discovery company based in Moscow, Russia specializing in small molecule screening and design through the use of its proprietary technology platform.
About CU-Boulder
As the flagship university of the state of Colorado, CU-Boulder is a dynamic community of scholars and learners. As one of 34 U.S. public institutions belonging to the prestigious Association of American Universities (AAU) – and the only member in the Rocky Mountain region – we have a proud tradition of academic excellence, with four Nobel laureates and more than 50 members of prestigious academic academies. CU-Boulder has blossomed in size and quality since we opened our doors in 1877 – attracting superb faculty, staff, and students and building strong programs in the sciences, engineering, business, law, arts, humanities, education, music, and many other disciplines. Today, with our sights set on becoming the standard for the great comprehensive public research universities of the new century, we strive to serve the people of Colorado and to engage with the world through excellence in our teaching, research, creative work, and service.
Friday, June 6, 2008
Docking validation study: PDK1-kinase (oncology)
Pyruvate dehydrogenase kinase, isozyme 1, also known as PDK1, is a human gene.It codes for an isozyme of pyruvate dehydrogenase kinase (PDK).Pyruvate dehydrogenase (PDH) is a part of a mitochondrial multienzyme complex that catalyzes the oxidative decarboxylation of pyruvate and is one of the major enzymes responsible for the regulation of homeostasis of carbohydrate fuels in mammals. The enzymatic activity is regulated by a phosphorylation/dephosphorylation cycle. Phosphorylation of PDH by a specific pyruvate dehydrogenase kinase (PDK) results in inactivation.
There are no as much known inhibitors as for thrombine. BindingDB gives a few more than 70 compounds with measured binding affinities, all relatively strong binders, many of them similar to each other. We run our QUANTUM software to perform docking and the affinity calculations. The results are represented on the graph and demonstrate a solid correlation. In fact the correlation shows QUANTUM's ability to identify strong binders and distinguish between similar compounds (selectivity).
Thursday, June 5, 2008
Docking validation study: classic example, thrombine
We have extracted the binding data from the BindingDB database and docked all the molecules onto a single (of a few available) 3D structure (2cn0 from the pdb databank).
The figure represents graphically the results of the research. The calculated and the measured activities are well correlated. Strong binders are indeed identified as strong binders (left bottom part of the graph). The accuracy of the predictions is quite good (see our discussion on the quality of the biological data here and here).
The results of the calculations can be conveniently summarized in terms of confidentiality matrix. Normally a first screen of novel compounds is performed at a certain concentration to distinguish between the active and non-active compounds. Let's take a standard, 1muM (~-35kJ/M) activity, as a separation cut-off. Then the confidence matrix has the following elements:
- Experimentally active, Predicted active: 29 molecules
- Experimentally n-active, Predicted active: 15 molecules (false positives)
- Experimentally active, Predicted n-active: 8 molecules (false negatives)
- Experimentally n-active, Predicted n-active: 156 molecules
Monday, May 26, 2008
Protein Flexibility and False Positives detection.
Let us show first that a calculation with flexible protein gives a reasonable prediction of the binding free energy. To do that we selected a set of ~200 protein - ligand complexes from the BindingDB database. The protein-ligand pairs were selected mainly so that the complex is small and therefore the whole calculation is fast. The results are represented on the Figure. The horizontal and the vertical axis represent the calculated and the experimental value of the binding free energy calculated from the complexed positions of the ligand within the protein.
The correlation is clearly there and in a few days I will show that the calculated values demonstrate not only the accuracy, but also a good selectivity.
The other Figure represents the correlation between the results of rigid receptor fast docking procedure (horizontal axis) and the fully flexible binding free energies (vertical axis). Although the rigid protein force field has a decent correlation, it fails to recognize electrostatic clashes and thus leads to a fairly large amount of false positives among the predicted ligands. Only about 10% of all the ligands, all originally predicted in the muM range survives as binders. The trend is also clear: all the binding energy values increase (fully flexible force field gives less binders than the rigid calculation would suggest).
Saturday, May 24, 2008
Quantum Pharmaceuticals and Tibotec Pharmaceuticals enter antiviral drug discovery collabortaion.
Under the terms of the agreement Quantum Pharmaceuticals will provide Tibotec Pharmaceuticals with the family of anti-viral drug hits. The drug hits were discovered by Quantum Pharmaceuticals using its proprietary discovery technology. Tibotec Pharmaceuticals is to further evaluate and develop transferred molecules. Financial terms of the collaboration were not disclosed.
About Tibotec Pharmaceuticals BVBA
Tibotec BVBA is a global pharmaceutical and research development company. The Company's main research and development facilities are in Mechelen, Belgium with offices in Yardley, Pa. and Cork, Ireland. Tibotec is dedicated to the discovery and development of innovative HIV/AIDS and hepatitis C drugs, and anti-infectives for diseases of high unmet medical need.
About Quantum Pharmaceuticals
Quantum Pharmaceuticals is a drug discovery company based in Moscow, Russia specializing in small molecule screening and design through the use of its proprietary technology platform.
Wednesday, May 21, 2008
Computer aided drug design video from Quantum Pharma
Molecular modelling software of Quantum Pharmaceuticals is used to dock small molecule to active site of target protein. The molecular docking on flexible protein is explored. The Quantum docking software is available for free use at LeadFinding.com, the online hit-to-lead optimization service to filter and profile chemical compounds in chemical database of ChemDiv - organic chemistry supplier.
Wednesday, April 9, 2008
Model of Intestinal Passive Absorption
hd and hm – are thickness of the diffusion layer adjacent to the intestinal wall and the enterocyte’s membrane;
Dw and DL are diffusion coefficients of the drug in the intestinal lumen (can be approximately described by the diffusion coefficient in water), and in the drug membrane. D – distribution coefficient of the drug.
The drug diffusion to apical membrane of enterocytes and through it is described by Fick’s law:
dJ1/dt = -Dw gradC = -Dw (C1-C0)/hd
dJ2/dt = -DL gradC = -DL C1D/hm.
(here we supposed the the blood flow is high that the drug concentration in the blood is zero)
In the steady-state
dJ1/dt = dJ2/dt = dJ/dt,
and
C1 = (Dw/hD)/ (DLD/hm + Dw/hd) * C0
Therefore apparent permeability of the drug is:
Papp = [dJ/dt] / ΔC = [dJ/dt] / C0 = (Dw/hd)/(1 + Dwhm/(hdDLD)) (*)
The Figure represents experimental LogPapp plotted against LogD for drugs that are reported not to be subjected to active transport, active efflux, and paracellular diffusion (blue points); and model predictions (solid line). RMSD = 0.34 log units.Tuesday, March 25, 2008
Quantum Pharmaceuticals and Otechestvennye Lekarstva OJSC entered drug discovery collaboration.
(Moscow, 25 March 2008) Quantum Pharmaceuticals entered a collaboration with Otechestvennye Lekarstva OJSC. Under the terms of collaborative agreement Quantum Pharmaceuticals will apply its proprietary technology to explore the mechanism of action and new therapeutic uses of several compounds of Otechestvennye Lekarstva OJSC. Quantum Pharmaceutical will utilize its proprietary protein panel representative for human proteome. The financial terms of the agreement were not disclosed.
About Quantum Pharmaceuticals
Quantum Pharmaceuticals is a drug discovery company based in Moscow, Russia specializing in small molecule screening and design through the use of its proprietary technology platform.
About Otechestvennye Lekarstva OJSC
Otechestvennye Lekarstva OJSC is one of the largest Russian pharmaceutical companies with recent turnover of US$150 million. The company has 5,000 employees and markets over 200 products in Russia and in 20 other countries. Otechestvennye Lekarstva OJSC was selected as the best pharmaceutical company in Russia in 2005 and 2006 by the Russian government.
Thursday, January 24, 2008
LD50 vs. MRDD: what's death for a mice is good enough for a man
Accurate prediction of such endpoints is only possible if both quantities are "physical" characteristics of a compound, rather than signatures of ever changing views of regulating agencies.
The plot on the left represents the "correlation" between experimental values of MRDD (according to FDA) and LD50 (rat) taken from different sources. As you can see, both quantities have a reasonable degree of correlation for low or intermediate toxicity levels. As soon as toxic compounds are considered, the correlation is lost and apparently no good prediction starting from physical properties of a molecule can be done.
For a moderately toxic molecule we can derive an approximate relation:
-LogMRDD = -LogLD50+2.
In "a human language": the lethal and the maximum recommended dose are roughly two orders of magnitude different; a concentration killing a mice is in fact the maximum recommended for a human being.
Friday, January 18, 2008
q-hERG: QUANTUM's innovative approach to hERG binding calculations is finally released
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.
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.
And continue to CHECK OUT |
Friday, January 11, 2008
HERG binding prediction quality: q-HERG model vs. experiments
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.
Wednesday, January 9, 2008
Drug likeness: what do bioavailability and toxicity properties tell us about druglikeness?
A good drug should show good availability, low toxicity and high potency. The quantitative measures of such properties are bioavailability (BA, measured in %), Maximum Recomended Daily Dose (MRDD, mmol/L) and IC50 against a drug's target.
The product of toxicity and availability, MRDD*BA, gives an upper bound on target IC50 and hence is an indication of a drug quality. The Figure above represents the distribution of such product for slightly over 100 drugs. As it can be seen from the Graph, most of drug compounds have the product small, roughly below 2*10^-5mol/L. Hence, small value of MRDD*BA product may be regarded as an indication of druglikeness.
In fact the situation gets even more interesting if the same druglikeness parameter is plotted in log-scale (see the Figure on the right). Since MRDD*BA limits drug's IC50 against its target, we can deduce that most drugs are centered around pIC50 = 5 (which means that the target pIC50 should exceed 5).