To What Extent can the Calculation of ADMET of Candidate Drugs be Predicted?

The discovery and development of new drugs is expensive & time-consuming! Whether a candidate can become a medicine is a work that R&D personnel continue to explore! When a molecule has obtained the desired drug effect, the work of preparing the drug begins to spread, and the research work of “absorption-distribution-metabolism-excretion-toxicity” (ie ADMET) is also carried out. Foreseeing the candidate varieties of ADMET in advance will undoubtedly greatly improve the efficiency of drug development. So, can calculations do this? To what extent can it assist?

01 Important Parameters Related to ADMET

In the final analysis, the ADMET of a drug depends on the physical and chemical properties and biological properties of the drug molecule itself; the most relevant indicators of these physical and chemical properties are: lipophilicity, hydrogen bonding, solubility, permeability, and so on.

➣ Lipophilic

The important indicator is LogP, which represents the ratio of the compound concentration when the oil phase and the water phase are in equilibrium. Lipophilicity is a physical and chemical parameter that must be considered when developing new drugs. It has a significant impact on pharmacokinetic properties, such as absorption, distribution, permeability, and drug clearance pathways. High fat solubility can often satisfy a certain selectivity and efficacy. For example, the targets of neurotransmitter pathways and some targets in cells usually need to be combined with lipophilic agonists to achieve the desired effect. Therefore, automatic calculation and measurement of water solubility, lipophilicity and ionization degree have been applied and integrated into the drug discovery and development phase.

➣ Hydrogen bond

Hydrogen bonding is considered to be the driving factor that has a significant effect on the biological activity of the compound, which reflects the interaction between the HB acceptor target and the donor compound, and vice versa; The permeability of the biofilm plays a role; and, in order for the compound to pass through the biofilm, the hydrogen bond must be broken in the water environment of the drug. A large number of studies have shown the relationship between hydrogen bonding and QSAR models. Therefore, quantifying the strength of hydrogen bonds is considered an important part of the drug design stage.

➣ Solubility

Intrinsic solubility can be defined as the thermodynamic solubility of a drug at pH. The amount of poorly soluble or low-soluble drugs into the blood circulation is small, so it cannot provide the necessary efficacy. In most cases, it is difficult to find molecules with satisfactory BCS classification. Therefore, calculations can also be used to predict solubility and improve drug absorption, which has further prompted scientists to pay close attention to the prediction of drug solubility during drug development in the past few years.

➣ Permeability

Osmotic drugs mainly cross the biological barrier including the intestinal epithelium and the BBB through passive diffusion mechanisms, in which substances are transported through the action of a concentration gradient. Most of the results show that the lipophilic distribution, molecular size, HB binding capacity of the drug can be used to predict the membrane permeability of the drug as accurately as possible, thereby saving the drug development cycle.

02 Calculate the Characteristics of Macro Forecasts in ADMET

ADMET prediction technology dates back to 1863. The earliest relevant is the impact of drug solubility prediction on potential adverse reactions. Later, the application of computing in predicting the pharmacokinetic properties of drugs has also received more and more attention.

➣ Absorption prediction

The absorption of oral drugs mainly includes promotion of diffusion, passive diffusion and active transportation; absorption capacity is determined by various parameters, and in some cases, it can be described by the permeability or solubility of the drug. In the calculation model, there are various models of in vivo and in vitro prediction technologies that are combined for the prediction of oral absorption. For example, permeability. Generally speaking, there is a certain relationship between the transmembrane and lipophilicity of drugs in the intestine. The absorption of drugs with low molecular weight or no significant metabolism is usually determined by whether the intestine can penetrate. Calculations can provide an assessment of intestinal permeability and are defined as fast and cheap. Basic models including PSA, rapid PSA and other complex models are used to predict intestinal permeability.

➣ Distribution forecast

The prediction of tissue distribution can promote the prediction of pharmacodynamics and toxicokinetics; distribution prediction is mainly related to BBB permeability, apparent volume of distribution (VD) and plasma protein binding (PPB), which are used in determining drug regimens, Both the plasma concentration and the penetration rate of the entire BBB have a significant effect, and are especially helpful in predicting central nervous system targets, side effects and non-central nervous system drug treatments. Such as drugs, PPB can affect the pharmacokinetics and efficacy of drugs; the most important protein involved in the binding of drugs in plasma is human serum albumin, which can bind to a variety of endogenous and exogenous molecules. There are many computational models that can predict the interaction of molecules with human serum albumin. Many of these models are based on the three-dimensional crystal structure of albumin, which can be used for docking studies to predict the binding of molecules to albumin. There are also QSPR models developed based on existing data on various ligands known to bind albumin.

➣ Metabolism prediction

Some research reports pointed out that compared with other pharmacokinetic parameters, the metabolism of drugs is the most difficult to predict parameter, because the metabolic process is a very complex process involving various enzyme activities and is different due to different genetic factors. Different calculation models are also being successfully applied to the prediction of some drug metabolism. Cytochrome P450 (CYP) is considered to be the most influential enzyme in drug metabolism, which promotes the development of many models, such as QSAR to predict the molecular metabolism of CYP enzymes. Another enzyme involved in drug metabolism is UGT enzyme, but compared with CYP, due to less data on the matrix and subtypes of UGT, the level of models predicting its molecular metabolism is lower.

➣ Excretion prediction

Although almost all drug substances are excreted from the body, researchers are seldom interested in the prediction of drug excretion parameters. Drugs in the second stage of metabolism usually exist in an unchanged form, usually lacking pharmacodynamic activity, and there is relatively greater interest in predicting the excretion of metabolites in this stage. After predicting drug excretion, the collected information must be integrated into the predictive model to provide a complete model to describe the substance behavior at different stages of drug discovery and development.

➣ Toxicity prediction

Toxicity assessment is a key issue that developers and researchers must pay close attention to. Computational toxicology, through the use of toxicity databases, makes QSAR modeling possible. Drug toxicity prediction can also reduce the need for animal testing and obtain more appropriate toxicity predictions. Toxicity prediction can predict systemic toxicity as well as the toxicity of a certain organ. The prediction of carcinogenicity and genotoxicity has also been increasingly applied to drug development.