Undrugable Targets: The Major Challenge for Modern Drug Discovery (Receptor.AI Report)
AI approaches to beat inrugability
LONDON, UK, August 14, 2022 /EINPresswire.com/ — The scariest word in modern drug discovery is “irreversible.” This rather vague term means that it is notoriously difficult if not impossible to find a drug that would act on a particular molecular target (usually the protein).
According to recent estimates, up to 85% of all human proteins are considered non-medicinal. In reality, the entire modern pharmaceutical industry works with a rather limited set of “good” target proteins known to interact well with drugs of a particular chemical nature. The vast majority of functionally important proteins known to be associated with disease are never used as drug targets.
The reasons for non-medication are diverse and range from the “unfortunate” molecular structure of the protein to its “bad luck” in a totally unscientific and mystical sense:
1. The protein may lack a well-defined binding pocket on its surface, which the drug molecule could access;
2. The binding pocket could be very shallow, forming only a few molecular interactions;
3. The protein can only function properly in a supramolecular complex with other proteins;
4. The structure of the binding pocket could be shared by several different proteins, which leads to the lack of drug selectivity;
5. In order to bind effectively to the target protein, all drugs may share the same structural motif, which appears to be toxic or otherwise problematic in terms of the whole organism;
6. Sometimes drugs for a particular target don’t pass clinical trials for many years, one after another, and no one wants to invest in another failure;
7. Finally, sometimes the reasons for not taking medication are simply unknown.
Most drug discovery companies attempt to target impossible-to-digest targets using a wide variety of techniques. Receptor.AI does not stay away from this trend. Recently, we started working on our own platform to target previously unmedicated proteins.
Our approach is pragmatic and based on our experience of “good” drug targets.
– We do not address all unrecoverable targets, but rather focus on the most common cases of irrevocability.
– We continue to work with the classic small molecule workflow (although the compounds may be less small in this case).
– We prioritize the ease and low cost of synthesis of our drug candidates despite the “weird” nature of their targets.
Multimeric protein complexes
The first class of non-drug targets we intend to address are multimeric protein complexes, which form the functional binding interface between their subunits. Proteins, which function in isolation, are mostly enzymes. All the rest of our proteome is involved in protein-protein interactions in one way or another, but not all of these interactions are drug-induced. We focus on stable protein-protein complexes, which form well-defined inter-protein interfaces. Such interfaces could be targeted by small molecule drugs in the same way as isolated protein binding pockets.
The first obstacle in the way of these multimeric targets is to obtain their 3D structure. We use X-ray crystallography, NMR or Cryo-EM structural data, homology modeling and de novo structure prediction using the famous and most accurate AlphaFold2 algorithm.
Once the structure of the protein assembly is determined, we optimize and balance it using molecular dynamics simulations.
Finally, the crucial AI-based step begins. Our proprietary ML model is able to recognize which chemical groups will bind favorably in different parts of the protein-protein interface. Another state-of-the-art deep learning model uses this information to perform sophisticated virtual screening coupled with on-the-fly molecular generation and assigns a binding score for each potential compound. Such AI scoring is extremely fast, allowing the screening of huge chemical spaces and the probing of large protein-protein interfaces.
The resulting compounds are guaranteed to be not only chemically correct but also easy to synthesize. The resulting compounds contain the chemical characteristics that allow them to bind strongly and selectively to a specific protein-protein interface.
Shallow binding pockets
Proteins with shallow binding pockets are common and impose a significant challenge for drug discovery. The number of possible interactions without binding between the drug and the pocket is very limited, which makes the binding not only weak but also less selective.
To overcome this problem, we have developed a technique that uses protein surface areas around the pocket to increase the number of useful molecular interactions. We specifically favor larger compounds, which extend outside of the pocket and additionally anchor around it. Despite the nonspecific nature of the binding outside the pocket, it helps hold the drug in place and allows weak interactions inside the pocket to facilitate selectivity.
In many cases, the protein-protein interfaces do not form any pocket slits, which could be targeted directly. Instead, the coupling of proteins in the assembly could be prevented from inhibiting their function. In such a case, the potential drug must mimic one of the partner proteins and bind to the other partner to prevent assembly.
However, peptides are not always the most desirable drugs due to their poor stability against proteases, high toxicity, suboptimal bioavailability, and problems crossing cell membranes. We use a slightly different approach. We plan to design rather large non-peptide compounds, which expose the same chemical groups as the partner protein. We intend to achieve this by matching the compounds to the spatial array of pharmacophores inferred from the surface of the partner protein. The resulting compound is not a peptide, which makes it stable against proteases and other metabolic transformations.
The problem of undrugable targets is too complex and too vast to be solved by a single technique. Different classes of non-drug targets require different approaches, and there is no guarantee that any non-drug target can be made drug-like. Nevertheless, even a small fraction of new drug targets can lead to huge advances in pharmacology and medicine and lead to cures for incurable diseases. Receptor.AI joins other companies in hunting the holy grail of tamper-evident targets using its unique technology stack and scientific expertise.