Artificial intelligence in cancer target identification and drug discovery

Abstract

Artificial intelligence is an advanced method to identify novel anticancer targets and discover novel drugs from biology networks because the networks can effectively preserve and quantify the interaction between components of cell systems underlying human diseases such as cancer. Here, we review and discuss how to employ artificial intelligence approaches to identify novel anticancer targets and discover drugs. First, we describe the scope of artificial intelligence biology analysis for novel anticancer target investigations. Second, we review and discuss the basic principles and theory of commonly used network-based and machine learning-based artificial intelligence algorithms. Finally, we showcase the applications of artificial intelligence approaches in cancer target identification and drug discovery. Taken together, the artificial intelligence models have provided us with a quantitative framework to study the relationship between network characteristics and cancer, thereby leading to the identification of potential anticancer targets and the discovery of novel drug candidates.

Fig. 1

The historical milestones of network-based and ML-based biology analysis. (Created with BioRender.com)

Fig. 2

Artificial intelligence to integrate multiomics data (e.g., epigenetics, genomics, proteomics, and metabolomics) for cancer therapeutic targets identification. (Created with BioRender.com)

Fig. 3

The flow chart of the shortest path algorithm. The red paths in the bottom network are the identified shortest path from node S to T

Fig. 4

The flow chart of the module detection algorithm

Fig. 5. Four types of node centralities of biological networks.

(a) Degree centrality; (b) Coreness centrality; (c) Betweenness centrality; (d) Eigenvector centrality

Fig. 6

An illustration of a simple decision tree model

Fig. 7. An example of an ADTree model.

The root nodes indicate the ratio between positive and negative class examples. The numbers in parentheses within each decision node (rectangles) indicate the order in which the rule was found. The amount of node conservation between each of the trees is indicated by the colour of the box. Ovals (prediction nodes) contain the value for the weighted vote. The numbers next to the arrows correspond to the threshold for the prediction

Fig. 8. The illustration of graph-based neural networks for ML-based biology analysis.

The graph-based neural networks take the topology of the biological networks data (such as gene-gene networks, protein-protein networks and drug-target networks) as input data. And then, the graph-based neural network realizes the functions of link prediction, classification and clustering by analyzing the biological information in the network topology. (Created with BioRender.com)

Fig. 9

The workflow to identify novel anticancer targets by network-based. (Created with BioRender.com)

Fig. 10

The workflow to evaluate the druggability of potential target proteins. (Created with BioRender.com)

Fig. 11

The graph-based neural network for DTI prediction by combining both bottom-up and top-down biology analysis approaches. (Created with BioRender.com)

Fig. 12

The graph-based neural network capture the features related to drug properties from drug molecular structure to predict ADMET properties of drugs. (Created with BioRender.com)