Edition 2 / May 2019
For some years now Artificial Intelligence (AI) has entered a new hopeful era, particularly due to the rise of machine learning. Made possible by new algorithms, multiplication of data sets and the increase in computing power, applications are being utilized in many areas such as translation, autonomous car control, cancer detection, etc. The development of AI is done in a marked technological context by datafication, which affects all domains and sectors, robotics, blockchain, supercomputing and massive storage. These different technological realities will surely pave the way for the future of Artificial Intelligence.
The global peptide API market is estimated to be valued at US $1,700 Million in 2018, and likely to expand at a CAGR of 7.7% over the forecast period (2018–2025) to reach US 2,900 Million by 2025.
The most emblematic AI tool is Watson Computer Software from the IBM industrial group, introduced in 2005 into the healthcare market. Watson has been used in particular at the Memorial Sloan Kettering Cancer Center, an American Institute which incorporates innovative collaborations in medical research for diagnostic assistance and therapeutic proposal of cancer treatment. This type of software, presented as a “smart” tool for medical decision support, synthesizes a wealth of information from millions of medical reports, patient records, clinical tests and knowledge from medical research. Some software may soon diagnose cancer as well, or better, than specialists. According to a recent study,1 Artificial Intelligence has been able to automatically detect breast cancer with a success rate of 92%, almost equivalent to what specialists (96%) can reach. When combined with the physician’s analysis and diagnostic methods from automated software, the success rate is 99.5%, with a greatly reduced error risk (Fig 1).
The average timeline for a drug to be developed – from discovery phase to market launch – is about 15 years, and while the average cost of developing a new drug has risen to over $2.5 billion, the rate of clinical trial failure for new drugs is about 90%. The pharmaceutical industry is reaching a critical stage, and companies are struggling to achieve their return on investments. New and efficient solutions are now necessary to significantly cut the costs of production. This is not only important for pharma companies but also for patients, as the prices of drugs are skyrocketing, particularly in the US.
With increasing adoption of Big Data, there is a striking increase in market research firms foreseeing huge gains in the predictive analytics market for healthcare. Moreover, those working with data in healthcare organizations perceive the advent of technology that fuels the future of patient care, quality control, and R&D. The healthcare industry is currently going through a major transformation with the increasing adoption of Big Data and advanced analytics.
The advances in AI hold real promise to increase efficiency and bring down drug prices. Highly developed AI programs now have the capacity to identify patterns from enormous sets of data, and to generate algorithms to explain them. The analysis of electronic medical records and public health data could lead to quick identification of potential molecular targets for a disease. This will further help in making faster and more accurate hypotheses, which will in turn make the drug discovery process less expensive and more efficient. There are currently over 100 companies that are applying AI algorithms and predictive analytics to healthcare.
Recently, T. J. O’Donnell and A. Rubinsteyn published an open-source Class I MHC Binding Affinity Prediction software package for peptide / MHC Class 1 in the vaccine design field to predict the binding affinity of major histocompatibility complex I.
Very active in the AI space, GSK signed a $43 M drug discovery collaboration with U.K.-based AI-driven startup Exscientia to identify molecules for selected targets across undisclosed therapeutic areas. Using a rapid “design-make-test” cycle, Exscientia is able to design new molecules using their AI-system, as well as phenotypic and high content screening data, to assess their potency, selectivity and binding affinity towards specific targets. The projects will be heavily supported by Exscientia’s Big-data resources from their medicinal chemistry and large-scale bio-assays.
A month earlier Takeda announced a multi-year research partnership with the AI-driven drug design company Numerate to develop new clinical candidates in oncology, gastroenterology, and central nervous system disorders. Numerate plans to apply AI-based modeling at every stage of the process from hit finding and expansion through lead design / optimization, absorption, distribution, metabolism, and excretion (ADME) / toxicity predictions. As stated on their website, Numerate’s AI-platform is able to work with data points obtained from different studies of high-content, low-throughput phenotypic assays as well as high-throughput screening, structure-based design and traditional computational methods. Trained with versatile information, the AI-system can probe very large chemical spaces and identify the most promising drug candidates.
Numerate is also involved in a recently announced partnership with Servier, focusing on the design of small molecule modulators of ryanodine receptor 2 (RyR2), a highly challenging target identified as important in cardiovascular diseases.
1 Wang D., Khosla A., Gargeya R., Irshad H., Beck A. H. (2016), Deep Learning for Identifying Metastatic Breast Cancer, Beth Israel Deaconess Medical Center (BIDMC) at Harvard Medical School.
Stay up to date with the latest thought leadership articles and news from CordenPharma.