Unlocking The Future: 4 Key Ai And Machine Learning Trends Reshaping Pharma

By Author: Giselle Fernandes
Total Articles: 23
Comment this article

Drug Discovery and Development using AI in Pharma
The time and cost to bring a new drug from concept to market is 7-10 years and costs $2 billion. source

Additionally, the process from discovery to market is often met with high failure rates at various stages in the drug development process. The collaboration between Cyclica & Bayer to accelerate drug discovery and the work of Verge genomics to simplify the clinical development process are just a few companies that have been focusing efforts in this sector.

Cyclica & Bayer
Cyclica, a biotechnology firm, partnered with Bayer to advance drug discovery through a holistic set of Artificial Intelligence & computational biophysics products by screening small molecule drugs against pre-existing repositories of proteins. The algorithms, working with protein and drug efficacy data, predict ADMET (Absorption, Distribution, Metabolism, Excretion, and Toxicity) (Lehrer, 2019) to drastically reduce discovery time.

Verge Genomics
Three pharmaceutical firms and several tissue banks have partnered with Verge Genomics, a neuroscience firm, to ...
... bring AI-generated compounds to the market to treat neurodegenerative diseases. Verge builds large databases source of patient tissues, capturing human disease complexity and through AI, and combines the data with human genetics to discover novel gene targets.Rare Diseases & Personalized Medicine
Pharma companies generally have focused their research efforts on diseases that affect large segments of the population. The regulatory compliance needs, long drug development times, expensive clinical trials, and the need to keep drugs affordable contributed to the industry’s focus on releasing a few blockbuster drugs.

In the last ten years, with the advent of AI in the pharmaceutical industry and machine learning, pharma companies have shown more interest in developing drugs for rare diseases (source). The partnership between Tencent Holdings and Huma showcases the power of AI in the detection and diagnosis of certain rare neurological diseases.

Additionally, algorithms are being used to develop personalized drug treatments based on an individual’s genes, environment, and lifestyle. Drugs are being personalized on composition, quantity, frequency, and method of administration. The joint effort between Aprecia and Cycle Pharmaceuticals focused on the development of orphan drugs.

Tencent Holdings and Huma
Technology conglomerate Tencent Holdings, in collaboration with healthcare firm Huma, is creating an AI platform to use motion capture technology to track patient movements. Deep learning, image recognition & video analysis are used both to evaluate the severity of a patient’s condition and significantly reduce motor test time.Aprecia and Cycle Pharmaceuticals
Aprecia manufactures the epilepsy drug, Spritam, that is easy to administer by caregivers. Combining the precision of 3D printing and formulation science, the technology produces a tablet that disintegrates with a sip of water in just a few seconds. The company later partnered with Cycle Pharmaceuticals (source), to grow and commercialize orphan drugs those greatly improved patient quality-of-life.

Clinical Trials
A CBInsights6 study showed that 80% of clinical trials fail to meet enrollment timelines and 1/3 of phase 3 trials terminate due to enrollment difficulties.

Patients find the national registry difficult to navigate and clinical trial staff must carefully search through patient health information to see if a candidate could be a fit. Two start-ups, Deep (download PDF) and Antidote, are solving the problems by using machine learning algorithms to parse, analyze, and interpret health data thus streamlining the patient selection process.

Deep6
Deep6’s AI software analyzes both structured and unstructured data such as ICD-10 codes, doctor’s notes, and medical reports. NLP algorithms extract new clinical data points from the data and researchers filter results for specific conditions and traits. The algorithm also makes predictions about the condition of a person based on extracted data, marking that person a potential candidate thus eliminating the enrollment difficulties for pharmaceutical firms.

Antidote
Antidote utilizes its AI platform to match patients and trials, rather than focusing on recruiting patients to a clinical trial. Working in conjunction with healthcare organizations such as JRDF and Lung Cancer Alliance, they serve people searching for a clinical trial based on their health conditions. Patients interested in a trial are surveyed on their medical history and based on the responses, Antidote’s algorithm categorizes source and identifies clinical trials most suitable for the patient.