Title:	In silico Design of Drugs and Diagnostic Agents for Various Neurodegenerative Diseases
DNr:	SNIC 2020/14-1
Project Type:	SNIC Small Storage
Principal Investigator:	Murugan Natarajan Arul <murugan@kth.se>
Affiliation:	Kungliga Tekniska högskolan
Duration:	2020-02-01 – 2021-02-01
Classification:	10407
Homepage:	https://www.kth.se/profile/murugan
Keywords:

Abstract

In this project, we aim to design drugs and diagnostic agents for various neurodegenerative diseases such as Alzheimer's and Parkinson's using multiscale modeling approaches. Since the average lifespan in Sweden is high, the old population tend develop various aging related diseases including Alzheimer's and Parkinson's. There are no treatment methods available for these diseases yet. Also the approaches for the early diagnosis of these diseases are also not developed. So, developing effective methods for diagnosis and therapeutics will help the healthcare burden in future. In this proposal, we aim to contribute to this subject through computational modeling. We will use an integrated approach involving molecular docking, molecular dynamics, ab initio molecular dynamics and hybrid QM/MM response calculations. In disease diagnostics or therapeutics, we target either one or another target biomolecules with the use of small molecules and these compounds should bind to the targets with high binding affinity and specificity. These properties in turn depend on the free energy of binding and can be estimated computationally using different approaches including molecular docking, implicit simulation free energy calculations approach and quantum mechanics fragmentation scheme. In order to rank the compounds as per experimental data, we need to estimate the binding free energies within a few kcal/mol accuracy which is a major challenge in computing. So along with studying various protein-ligand interactions, a major part of the project also deals with the method development for the reliable estimation of binding free energies. Once the methods are validated by testing against different available datasets of binding affinities, they can be used for prediction purpose. In particular, here the targets associated with various neurodegenerative diseases will be mainly focused. Below we briefly discuss the relevant targets and mechanism behind diagnosis of the neurodegenerative diseases. In general, diseases in human-beings can be diagnosed by monitoring the aberrantly expressed biomolecules. In the case of Alzheimer's disease, monoamine oxidase-B (MAO-B), Acetyl choline esterase (AChE), Beta-secretase (BACE), amyloid and tau fibrils are suitable biomarkers and by designing small molecules which can specifically bind to these targets, we can easily diagnose diseases. In the design of diagnostic agents, off-target binding is the major problem. Since there are may other biomolecular structures co-exist in brain, computational approaches are very useful to screen the compounds for off-target binding. We compute the binding free energies using different approaches for compounds against aforemenationed targets. We benchmark these methods to reproduce the available binding affinity data. Once the methods are effective, we can use them for establishing structure-property (binding affinity) relationship. With this new diagnostic agents with improved binding affinity and binding specificity can be developed. Currently, we have studied the small molecules interaction with various targets such as BACE, MAO-B, amyloid and tau fibril and QM fragmentation and MM-GBSA approaches were found useful for ranking compounds. So, these approaches will be used for designing novel small molecular compounds for drug/diagnostic purpose for these targets.