How AI Is Laying The Foundation In The Discovery Of Drugs?

Fri, 24 January 2020 1:55

Drug discovery is the process in which new medicines are identified for treating or curing human diseases. The traditional way of drug discovery involved extracting ingredients from natural products and then conducting fundamental research to discover potential medicines. Historically, the speed of discovering new medicines was slow, baffling, and labor-intensive.

Accelerating the Process of Drug Discovery

Most of the drugs during the twentieth century were discovered by chemically synthesizing them, which still make up 90 percent of medications available today. Their benefits consist of simple manufacturing and administration routes. They likewise have a stable shelf life and low specificity, which means they are clinically safe and viable for large audiences. Nevertheless, low specificity can likewise result in side-effects, decreasing the odds of success in clinical preliminaries.

Since the 1990s, scientific and technological advancements have prompted the discovery of more complex, larger, organic therapeutics known as biologics, which are exceptionally specific to their target. Biologics have raised the interests of media and investors because of their novel techniques and potential to cure illnesses that were previously considered untreatable. The Food and Drug Administration (FDA) distinguished 17 out of 59 approved drugs as biologics.

Until now, modern drug discovery remains a lengthy and costly process, with high unsuccessful rates. The average launch time of a molecule is around 10 to 12 years. In 2018, Deloitte released a report titled, “Measuring the return from pharmaceutical innovation” that estimated the average cost of R&D for the top 12 biopharmaceutical companies to be $2.168 billion per drug. This cost is double to that of the $1.188 billion calculated in 2010. On the other hand, the statistic showed that the average peak forecast sales per new pharmaceutical pipeline asset dropped from $816 million in 2010 to $407 million in 2018 – which stands at almost half of the value derived in 2010. The resultant ROI decline in 2018 stood at 1.9 percent from 10.1 percent in 2010. Discovering methods for improving the efficiency and cost-effectiveness of introducing new medications to the market is critical for the business.

One way this could be achieved is by accelerating the speed of drug discovery and improving accuracy and predictability, which presently makes up for 1/3rd of the above costs. In all, 10,000 molecules were screened out of which, only 10 made it to the clinical trials. Besides, the success rate of a compound entering the phase 1 trials and phase 1 of clinical testing currently stands below 10 percent and has not increased during the last 10 years. Given the increased costs related to introducing a drug in the market, improvement in the accuracy of predictions by even 10 percent could save a lot of money spent on drug development.

Several AI-powered solutions are coming around which are significant for accelerating the process of drug discovery. While these are centered on changing the research process of small molecules, they are likewise showing potential in the identification of new biologics, for example, therapeutic antibodies against fibrosis, cancer, and other diseases. The capability of AI to improve the understanding of structures and specificity to the target molecules is because of the huge-scale availability of structured and unstructured scientific data.

Key Factors That Biopharmaceuticals Must Consider Before Adopting AI

AI learning algorithms can effectively extract features, concepts, and relationships from data and learn self-reliably from data patterns, supplementing what individuals do. The technology also likewise helps cross-reference published scientific content with substitute information sources, such as public databanks, clinical trial information, unpublished datasets, and conference abstracts. By implementing data mining, AI applications in the process of drug discovery have already introduced new potential candidate drugs, in a few months rather than years. If AI solutions are implemented at the drug-discovery stage, they have the potential to boost the efficiency of the R&D process.

In order to do this, biopharmaceutical companies need to seriously consider implementing a robust strategy to integrate AI solutions into their processes. And during the last three years, biopharma companies have adopted AI solutions to fuel their traditional drug discovery processes, including investing in startups, collaborating with tech giants or research institutions, and including AI experts and data analysts in their workforce. According to the Deep Knowledge Analytics (DKA) Landscape of AI for Drug Discovery and Advanced R&D Q2 2019 report, around 400 investors, 170 AI companies, and 50 corporations have made significant progress in the drug-discovery landscape.

'4P' Medicine

AI and other cutting-edge technologies opting for multiple information sources can improve the accuracy rate of targeted treatments and will help drive the health ecosystem towards a future where drug delivery is customized, predictive, participatory, and preventive. This will likewise result in new, increasingly efficient, and effective models of care. In the next 10 years, these driving forces will significantly affect treatments and patient outcomes, especially in regions of neglected care.

With the increasing number of AI-identified compounds, drugs fit for treating explicit pathologies will become available. By 2030, the process of drug discovery will be largely done by using simulation and in collaboration with research institutions and centers. Screening to preclinical testing processes will be implemented in just a few months rather than years and new potential candidates will be identified at much lower costs.

Major advances in the processes used for drug discovery will improve to lay the foundation for precision medicine to become standard. In the next ten years, patients can expect these improvements to have a key impact on the viability of their treatment alternatives and on disease outcomes, especially in regions right now without any treatments available.