Article
Engineering, Multidisciplinary
Joshua Bongard, Michael Levin
Summary: The applicability of computational models to the biological world is being debated, with the suggestion to adopt an observer-dependent view rather than strict boundaries between categories. Living systems are polycomputing, simultaneously performing multiple functions in the same place, which challenges prediction and control. Understanding and harnessing polycomputing can have significant impacts on fields like regenerative medicine, robotics, and computer engineering.
Review
Physics, Mathematical
E. Weinan, Huan Lei, Pinchen Xie, Linfeng Zhang
Summary: Neural network-based machine learning enables efficient and accurate approximation of high-dimensional functions, allowing for applications in multi-scale modeling in various fields such as molecular dynamics and non-Newtonian fluid dynamics.
JOURNAL OF MATHEMATICAL PHYSICS
(2023)
Review
Biochemical Research Methods
Thomas Gaudelet, Ben Day, Arian R. Jamasb, Jyothish Soman, Cristian Regep, Gertrude Liu, Jeremy B. R. Hayter, Richard Vickers, Charles Roberts, Jian Tang, David Roblin, Tom L. Blundell, Michael M. Bronstein, Jake P. Taylor-King
Summary: Graph machine learning (GML) is gaining attention in the pharmaceutical and biotechnology industries for its ability to model biomolecular structures and integrate multi-omic datasets. While still emerging, milestones such as repurposed drugs entering in vivo studies indicate that GML will become a preferred modeling framework in biomedical machine learning.
BRIEFINGS IN BIOINFORMATICS
(2021)
Review
Multidisciplinary Sciences
Telmah Lluka, Jonathan M. Stokes
Summary: As the global burden of antibiotic resistance continues to increase, creative approaches are needed to accelerate the development of new medicines. The rapid progress of artificial intelligence offers a promising solution by addressing bottlenecks in the antibiotic discovery pipeline. This review discusses the application of artificial intelligence in various areas of antibiotic discovery and proposes open access to high-quality screening datasets and interdisciplinary collaboration to speed up the development of new antibiotics.
ANNALS OF THE NEW YORK ACADEMY OF SCIENCES
(2023)
News Item
Multidisciplinary Sciences
Carrie Arnold
Summary: There are lingering questions regarding the ability of AI tools to truly disrupt the pharmaceutical industry.
Review
Pharmacology & Pharmacy
Ryan K. Tan, Yang Liu, Lei Xie
Summary: This survey reviews the state-of-the-art reinforcement learning methods and their applications in drug design. The challenges of applying reinforcement learning in systems pharmacology and personalized medicine are discussed, along with potential solutions.
EXPERT OPINION ON DRUG DISCOVERY
(2022)
Review
Pharmacology & Pharmacy
R. S. K. Vijayan, Jan Kihlberg, Jason B. Cross, Vasanthanathan Poongavanam
Summary: Artificial intelligence is playing a crucial role in drug discovery, from target identification to preclinical development. This review provides an overview of current AI technologies and presents real impact examples, while discussing the opportunities and challenges of adopting AI in drug discovery.
DRUG DISCOVERY TODAY
(2022)
Review
Biochemical Research Methods
Anna Torkamannia, Yadollah Omidi, Reza Ferdousi
Summary: Combination pharmacotherapy with synergistic effect is an effective strategy for treating complex diseases, but testing different compound combinations is expensive and challenging. Computational methods, particularly machine learning algorithms, have achieved significant success in predicting synergistic drug combinations in cancer.
BRIEFINGS IN BIOINFORMATICS
(2022)
Article
Food Science & Technology
Ning Tang
Summary: In this study, structure-based machine learning modeling and sweetness recognition mechanism were used to investigate the search for novel, natural, high-sweetness, low-calorie sweeteners. It was found that molecular connectivity, composition, tpsaEfficiency, structural complexity, and shape were closely related to the sweetness of a compound. The machine learning models showed very good performance in classifying sweet/non-sweet compounds and predicting their relative sweetness. Additionally, a specific binding pocket was identified for sweet compounds, providing useful information for developing new sweeteners.
Article
Oncology
Antonio Federico, Michele Fratello, Giovanni Scala, Lena Mobus, Alisa Pavel, Giusy del Giudice, Michele Ceccarelli, Valerio Costa, Alfredo Ciccodicola, Vittorio Fortino, Angela Serra, Dario Greco
Summary: The study found that current treatments for complex diseases have high toxicity and often lead to drug resistance, highlighting the need for novel and more specific pharmacological therapies. The authors developed an integrated network pharmacology framework combining mechanistic and chemocentric approaches to predict potential drug combinations for cancer therapy. The results show paclitaxel as a suitable drug for combination therapy in many cancer types and identified several non-cancer-related genes as potential candidates for cancer pharmacological treatment.
Article
Oncology
Ze-Jia Cui, Min Gao, Yuan Quan, Bo-Min Lv, Xin-Yu Tong, Teng-Fei Dai, Xiong-Hui Zhou, Hong-Yu Zhang
Summary: Breast cancer is a common disease with various subtypes, and the precision drug discovery strategy proposed in this study effectively identifies important disease-related genes in individuals and special groups, supporting its efficiency, high reliability, and practical application value in drug discovery.
Article
Biochemistry & Molecular Biology
Kaiyang Liu, Xi Chen, Yue Ren, Chaoqun Liu, Tianyi Lv, Ya'nan Liu, Yanling Zhang
Summary: Polypharmacology has emerged as a new paradigm in drug discovery, playing a crucial role in addressing polygenic diseases. This paper introduces multi-target-based polypharmacology prediction (mTPP), an approach that employs virtual screening and machine learning to explore the relationship between the action of multiple targets and the overall efficacy of drugs. Through the mTPP model, potential hepatoprotective components and candidates with potential effects against drug-induced liver injury (DILI) are identified. The model demonstrates accuracy in predicting the viabilities of APAP-induced injury cells, indicating its potential for aiding the development of polypharmacology and the discovery of multi-target drugs.
CHEMICO-BIOLOGICAL INTERACTIONS
(2022)
Review
Physiology
Karim Azer, Chanchala D. Kaddi, Jeffrey S. Barrett, Jane P. F. Bai, Sean T. McQuade, Nathaniel J. Merrill, Benedetto Piccoli, Susana Neves-Zaph, Luca Marchetti, Rosario Lombardo, Silvia Parolo, Selva Rupa Christinal Immanuel, Nitin S. Baliga
Summary: This paper reviews the history of mathematical biology and pharmacology models, highlights some gaps and challenges in the development and application of systems pharmacology models, and proposes an integrated strategy to overcome these challenges by leveraging advances in adjacent fields.
FRONTIERS IN PHYSIOLOGY
(2021)
Editorial Material
Multidisciplinary Sciences
Marissa Mock, Suzanne Edavettal, Christopher Langmead, Alan Russell
Summary: Artificial intelligence tools can facilitate data sharing on drug candidates while ensuring the security of sensitive information, enabling the utilization of machine learning and cutting-edge lab techniques for the benefit of society.
Review
Biochemistry & Molecular Biology
Sha Zhu, Qifeng Bai, Lanqing Li, Tingyang Xu
Summary: Drug repositioning plays a significant role in drug development and machine learning methods can accelerate this process. This article focuses on the repurposing potential of type 2 diabetes mellitus drugs for various diseases.
COMPUTATIONAL AND STRUCTURAL BIOTECHNOLOGY JOURNAL
(2022)