Exciting Possibilities ML Offers in Drug Discovery

Pharma Tech Outlook: Pharma Tech Magazine

Exciting Possibilities ML Offers in Drug Discovery

Pharma Tech Outlook | Wednesday, June 16, 2021

ML can help scientists and accelerate the drug discovery pathway.

FREMONT, CA: The conventional path of drug development is lengthy, expensive, and suffers from increased failure rates – scientists test millions of molecules. However, only a few progress to preclinical or clinical testing.  Embracing innovation, especially automated technologies, is critical to reducing the complexity connected with drug discovery and circumvents the high cost and time spent bringing medicine to market. Incorporating automation can make this hunt for drugs cheaper, effective, and less time-consuming. The last few years have witnessed significant growth in the use of new approaches and technologies in drug discovery. Here is more to know.

The availability of vast data sets and advanced algorithms has driven more interest and improvements in artificial intelligence (AI) in the field. AI can offer substantial enhancements at many stages of drug development, reducing the time from target identification to trials.  Machine learning, a subset of AI, is a quickly evolving field and is increasingly leveraged by many pharmaceutical companies. Integrating ML methods into the drug development process can aid in automating repetitive data processing and analysis tasks.

ML solutions are powered by big data modeling and analysis. The data can come from various sources and vary in a format making aggregating, storing, and preparing the data for analysis challenging, albeit needed.  ML trains a system to make decisions autonomously without any extra support. The decisions are made when the system learns and betters from experience –it learns from the data it had been offered and deciphered the associated patterns. ML tasks fall broadly into supervised learning, unsupervised learning, and sequential learning.

A novel deep-learning computer model created by researchers helps to predict correlations between gene expression and drug response. Leveraging the model, the team has found ten drug repurposing candidates for COVID-19. Two drugs have received regulatory approval; the remaining eight are presently investigational and tested in various indications.

Weekly Brief

Read Also