Drug development is recognized widely as a long, expensive and accidental process. Thousands of drug candidates have to undergo a series of tests and trials, and perhaps only one of them may turn out to be successful in the end. What’s worse, the cost of research and development of each new drug is about 2.558 billion US dollars, according to the Tufts Center for the Study of Drug Development.
The high cost and low success rate have hindered the development of new drugs and raised the threshold for new drug development. Many scholars have been making unremitting efforts to change the dilemma of drug research and development, but the slow development process of new drug R&D can still not be reversed. Now, the continuous advance of artificial intelligence (AI) technology seems to bring a new direction for the development of new drugs, perhaps it can at least change this bleak status quo.
Then how can AI actually facilitate new drug development and improve the whole efficiency? Let’s have a look.
1. AI can be used to screen biomarkers or targets
Many large pharmaceutical companies have employed AI in the screening of biomarkers or targets, which has become one of the key research directions.
In the case of Numedii, the researchers analyzed hundreds of millions of standardized, annotated biological, pharmacological, and clinical data through AI to obtain candidate drugs and biomarkers.
Recent studies have shown that anti-depressants found by NuMedii's technology are effective in a small cell lung cancer test model. Once new indications are identified and confirmed in appropriate preclinical models, NuMedii will also optimize the formulation and dosing for new uses of the drug to advance the project into early clinical stages.
2. AI can be used to construct new drug molecules
Admittedly, it is no easy task to construct new drug molecules with AI, and different companies have different purposes in constructing drugs molecules. Some companies have tried to use AI technology to help find similar chemical structures that are not protected by patents, so as to accelerate the development of generic drugs, while Insilico Medicine studies the structure of target biomacromolecules for drug molecular design.
The company's GANs platform uses a competitive neural network model to create new data that is different from real data, thereby dramatically reducing the time and cost of finding potential drug properties.
3. AI can be used to test the effectiveness and side effects of new drugs
Taking advantage of this, some companies offer drug candidate prediction services to pharmaceutical companies, startups and research institutions. Molplex has developed an AI technology platform named Optiplex, which can extract the link between disease and compounds from big data, predict the effectiveness and side effects of potential drugs, and help select the best drug candidate. In the United States, Atomwise used only one week to simulate two compounds for Ebola treatment.
4. AI can be used to develop new drug molecules
There are many ways to develop new drug molecules. The core is to use NLP algorithms to scan a large number of chemical libraries, medical databases and scientific papers published in conventional ways to identify novel drugs, drug genes and other treatment-related links, and to find potential new drug molecules.
BenevolentAI uses AI to understand and analyze a large number of bioscience materials through deep learning and natural language processing, including patent, genomic data, biomedical journals as well as over 10, 000 publications uploaded daily. Now a number of new drugs in clinical stages are found and an exclusive license and patents are also granted.
5. AI can be used for genetic analysis
For single-structured, massive genetic data, AI can effectively extract valuable information from the data, which is mission impossible for humans.
Engine Biosciences uses artificial intelligence to understand and test gene interactions, analyze generated data, decipher complex biological networks, test treatments for these interactions, and analyze and predict precision medical applications. Envisagenics helps researchers identify genes that are affected by alternative splicing (including those found in cancer and genetic diseases) through AI.
6. AI can be used for new drug targets and combination therapy
Watson aims to help life scientists discover new drug targets and alternative drugs. It can help researchers view different data sets and discover new links between drugs and diseases through dynamic visualization.
Watson's supercomputing power can be used in the development of new anticancer drugs via analyzing a large amount of publicly available data and the company's own data, constantly assuming drug targets as well as interacting in real time to obtain evidenced results in the end.
Watson's technology is mainly used for the following areas: the discovery of new drug targets in the field of immuno-oncology, the study of combination therapy, and the treatment strategies of patients. IBM Watson Health and Pfizer have signed an agreement to accelerate the development of new anticancer drugs, and compounds obtained have been used in the clinical trials for the treatment of Parkinson's disease.
About the author
Starting and specialized as a chemical supplier of inhibitors, APIs, metabolites and impurities since 2005, BOC Sciences now directs some focus on drug discovery via offering a wide range of related services, including virtual screening, antibody drug conjugate service, drug design, screening libraries, compound screening platform, formulation service, isotope labeling service, etc.