Drug Discovery: New Concepts Based on Machine Learning

A special issue of Pharmaceutics (ISSN 1999-4923). This special issue belongs to the section "Pharmaceutical Technology, Manufacturing and Devices".

Deadline for manuscript submissions: closed (1 April 2023) | Viewed by 4671

Special Issue Editors

Department of Chemistry, University of Fribourg, Chemin du Musée 9, 1700 Fribourg, Switzerland
Interests: drugs; computer chemistry; machine learning
Special Issues, Collections and Topics in MDPI journals
Department of Chemical, Paper, and Biomedical Engineering, Miami University, 64 Engineering Building 650 E High Street, Oxford, OH 45056, USA
Interests: thermodynamics; phase-equilibrium; molecular simulation; separation processes
Special Issues, Collections and Topics in MDPI journals

Special Issue Information

Dear Colleagues,

The ML/AI-based methodology era enables and opens new avenues that can boost the growth of new methods and their increasing importance. The computational-aided drug design exposes the impacts on drug discovery (new targets, the targeting of small molecules, targeted protein–protein interactions, SAR generation using data-driven experimental databases and integrated platforms, drug delivery pathways, etc.).

The Special Issue will cover the following topics:

  • targeting small molecules;
  • protein–protein interactions;
  • protein dynamics;
  • docking studies;
  • logP and pKa computational methods;
  • solvation-free energy;
  • QSPR/QSAR studies;
  • fragment-based drug discovery (FBDD).

We also welcome papers dedicated to computational and machine learning for drug discovery. The new ML approach for drug design and CADD was developed and designed for de novo drug design methods to generate a space for novel chemical compounds with desirable properties in a cost-efficient manner. This collection of articles highlights current developments in molecular generative models combined with machine learning and stresses the future directions for de novo drug design in combination with ML. We are happy to welcome papers dedicated to fragment-based drug discovery (FBDD) as a powerful tool to recognize and classify a new compound as the initial point for drug development.

Dr. Miroslava Nedyalkova
Dr. Andrew S. Paluch
Guest Editors

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 100 words) can be sent to the Editorial Office for announcement on this website.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Pharmaceutics is an international peer-reviewed open access monthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 2900 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.

Keywords

  • machine learning
  • targeting small molecules
  • protein–protein interactions
  • protein dynamic
  • docking studies
  • molecular dynamics
  • separation processes
  • solubility
  • activity coefficient
  • solvation-free energy
  • drug discovery

Published Papers (2 papers)

Order results
Result details
Select all
Export citation of selected articles as:

Research

19 pages, 3904 KiB  
Article
Protein Fitness Prediction Is Impacted by the Interplay of Language Models, Ensemble Learning, and Sampling Methods
Pharmaceutics 2023, 15(5), 1337; https://doi.org/10.3390/pharmaceutics15051337 - 25 Apr 2023
Cited by 3 | Viewed by 1865
Abstract
Advances in machine learning (ML) and the availability of protein sequences via high-throughput sequencing techniques have transformed the ability to design novel diagnostic and therapeutic proteins. ML allows protein engineers to capture complex trends hidden within protein sequences that would otherwise be difficult [...] Read more.
Advances in machine learning (ML) and the availability of protein sequences via high-throughput sequencing techniques have transformed the ability to design novel diagnostic and therapeutic proteins. ML allows protein engineers to capture complex trends hidden within protein sequences that would otherwise be difficult to identify in the context of the immense and rugged protein fitness landscape. Despite this potential, there persists a need for guidance during the training and evaluation of ML methods over sequencing data. Two key challenges for training discriminative models and evaluating their performance include handling severely imbalanced datasets (e.g., few high-fitness proteins among an abundance of non-functional proteins) and selecting appropriate protein sequence representations (numerical encodings). Here, we present a framework for applying ML over assay-labeled datasets to elucidate the capacity of sampling techniques and protein encoding methods to improve binding affinity and thermal stability prediction tasks. For protein sequence representations, we incorporate two widely used methods (One-Hot encoding and physiochemical encoding) and two language-based methods (next-token prediction, UniRep; masked-token prediction, ESM). Elaboration on performance is provided over protein fitness, protein size, and sampling techniques. In addition, an ensemble of protein representation methods is generated to discover the contribution of distinct representations and improve the final prediction score. We then implement multiple criteria decision analysis (MCDA; TOPSIS with entropy weighting), using multiple metrics well-suited for imbalanced data, to ensure statistical rigor in ranking our methods. Within the context of these datasets, the synthetic minority oversampling technique (SMOTE) outperformed undersampling while encoding sequences with One-Hot, UniRep, and ESM representations. Moreover, ensemble learning increased the predictive performance of the affinity-based dataset by 4% compared to the best single-encoding candidate (F1-score = 97%), while ESM alone was rigorous enough in stability prediction (F1-score = 92%). Full article
(This article belongs to the Special Issue Drug Discovery: New Concepts Based on Machine Learning)
Show Figures

Figure 1

15 pages, 4070 KiB  
Article
Prediction of α-Glucosidase Inhibitory Activity of LC-ESI-TQ-MS/MS-Identified Compounds from Tradescantia pallida Leaves
Pharmaceutics 2022, 14(12), 2578; https://doi.org/10.3390/pharmaceutics14122578 - 23 Nov 2022
Cited by 8 | Viewed by 1702
Abstract
Diabetes is a chronic disease that leads to abnormal carbohydrate digestion and hyperglycemia. The long-term use of marketed drugs results in secondary infections and side effects that demand safe and natural substitutes for synthetic drugs. The objective of this study is to evaluate [...] Read more.
Diabetes is a chronic disease that leads to abnormal carbohydrate digestion and hyperglycemia. The long-term use of marketed drugs results in secondary infections and side effects that demand safe and natural substitutes for synthetic drugs. The objective of this study is to evaluate the antidiabetic potential of compounds from the leaves of Tradescantia pallida. Thirteen phenolic compounds were identified from the ethyl acetate fraction of leaves of Tradescantia pallida using liquid chromatography-mass spectrometry. The compounds were then studied for the type of interactions between polyphenols and human α-glucosidase protein using molecular docking analysis. Prime Molecular Mechanics/Generalized Born Surface Area (MM-GBSA) calculations were performed to measure the binding free energies responsible for the formation of ligand–protein complexes. The compounds were further investigated for the thermodynamic constraints under a specified biological environment using molecular dynamic simulations. The flexibility of the ligand–protein systems was verified by Root Mean Square Deviation (RMSD), Root Mean Square Fluctuation (RMSF) and molecular interactions. The results authenticated the antidiabetic potential of polyphenols identified from the leaves of Tradescantia pallida. Our investigations could be helpful in the design of safe antidiabetic agents, but further in vitro and in vivo investigations are required. Full article
(This article belongs to the Special Issue Drug Discovery: New Concepts Based on Machine Learning)
Show Figures

Graphical abstract

Back to TopTop