ABSTRACT
Introduction
The transition from conventional cytotoxic chemotherapy to targeted cancer therapy with small-molecule anticancer drugs has enhanced treatment outcomes. This approach, which now dominates cancer treatment, has its advantages. Despite the regulatory approval of several targeted molecules for clinical use, challenges such as low response rates and drug resistance still persist. Conventional drug discovery methods are costly and time-consuming, necessitating more efficient approaches. The rise of artificial intelligence (AI) and access to large-scale datasets have revolutionized the field of small-molecule cancer drug discovery. Machine learning (ML), particularly deep learning (DL) techniques, enables the rapid identification and development of novel anticancer agents by analyzing vast amounts of genomic, proteomic, and imaging data to uncover hidden patterns and relationships.
Area covered
In this review, the authors explore the important landmarks in the history of AI-driven drug discovery. They also highlight various applications in small-molecule cancer drug discovery, outline the challenges faced, and provide insights for future research.
Expert opinion
The advent of big data has allowed AI to penetrate and enable innovations in almost every stage of medicine discovery, transforming the landscape of oncology research through the development of state-of-the-art algorithms and models. Despite challenges in data quality, model interpretability, and technical limitations, advancements promise breakthroughs in personalized and precision oncology, revolutionizing future cancer management.
Article highlights
Traditional drug discovery faces challenges in analyzing extensive datasets and unveiling hidden patterns efficiently, but AI and ML, especially DL, are revolutionizing this process by leveraging advanced computational techniques to accelerate the development of new therapeutic agents.
Key milestones in AI-driven drug discovery, such as GENTRL and AlphaFold3, illustrate advancements in clinical development and precision in drug discovery, though challenges like data bias and the need for high-quality datasets remain.
AI applications, leveraging techniques like classification, neural networks, and clustering, integrate complex multi-omics data to identify novel and reliable therapeutic targets, enhancing the understanding of carcinogenesis.
AI-driven generative models, including RNNs, GANs, and VAEs, autonomously learn molecular characteristics and create innovative compounds, enabling rapid exploration of wider chemical space.
AI-driven techniques in pharmacological prediction, preclinical applications, and drug repurposing are crucial for predicting DTIs, physicochemical properties, and ADMET characteristics, and discovering novel therapeutic strategies, thereby streamlining the drug development process.The future of AI-driven cancer drug discovery holds promise for revolutionizing oncology management through the integration of AI tools in every step of drug discovery, personalized medicine, and prognostic prediction, ultimately democratizing drug development and improving patient outcomes.
Abbreviation
AAE | = | Adversarial Autoencoder |
AEs | = | Autoencoders |
ADMET | = | Absorption, Distribution, Metabolism, Excretion, and Toxicity |
AI | = | Artificial Intelligence |
ANN | = | Artificial Neural Network |
AR | = | Androgen Receptor |
ATNC | = | Adversarial Threshold Neural Computer |
BRAF | = | B-Raf Proto-Oncogene |
CDK | = | Cyclin-Dependent Kinase |
CRISPR | = | Clustered Regularly Interspaced Short Palindromic Repeats |
CNNs | = | Convolutional Neural Networks |
CVAE | = | Conditional Variational Autoencoder |
DDR1 | = | Discoidin Domain Receptor1 |
DTA | = | Drug-Target Affinity |
DTI | = | Drug-Target Interaction |
DL | = | Deep Learning |
DNNs | = | Deep Neural Networks |
druGAN | = | drug-Generative Adversarial Network |
EGFR | = | Epidermal Growth Factor Receptor |
ERBB2 | = | Erb-B2 Receptor Tyrosine Kinase 2 |
FDA | = | Food and Drug Administration |
FP | = | Fingerprints |
GANs | = | Generative Adversarial Networks |
GCPN | = | Graph Convolutional Policy Network |
GCNs | = | Graph Convolution Networks |
GENTRL | = | Generative Tensorial Reinforcement Learning |
GNNs | = | Graph Neural Networks |
GPT | = | Generative Pre-Training |
HCC | = | Hepatocellular Carcinoma |
HTS | = | High-Throughput Screening |
kNN | = | k-Nearest Neighbour |
LapRLS | = | Laplacian Regularized Least-Squares |
LBVS | = | Ligand-Based Virtual Screening |
Log P | = | Partition Coefficient |
MANTRA | = | Mode of Action by Network Analysis |
ML | = | Machine Learning |
MLP | = | Multilayer Perceptron |
MoA | = | Mechanisms of Action |
MOGONET | = | Multi-Omics Graph cOnvolutional Networks |
MolGAN | = | Molecular GAN |
ORGAN | = | Objective-Reinforced Generative Adversarial Networks |
ORGANIC | = | Objective-Reinforced Generative Adversarial Network for Inverse-Design Chemistry |
PINNs | = | Pairwise Input Neural Networks |
PK | = | Pharmacokinetic |
QSAR | = | Quantitative Structure-Activity Relationship |
RANC | = | Reinforced Adversarial Neural Computer |
ReLeaSE | = | Reinforcement Learning for Structural Evolution |
RF | = | Random Forest |
RL | = | Reinforcement Learning |
RNNs | = | Recurrent Neural Networks |
SBVS | = | Structure-Based Virtual Screening |
SMILES | = | Simplified Molecule-Input Line-Entry System |
SVM | = | Support Vector Machine |
VAE | = | Variational Autoencoder |
VS | = | Virtual Screening |
2D | = | Two-Dimensional |
3D | = | Three-Dimensional |
Declaration of interest
The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.
Reviewer disclosures
One review is an employee of Charles River Laboratories. Peer reviewers on this manuscript have no other relevant financial or other relationships to disclose.