The future of medicine: AI’s role in uncovering new drugs

Artificial intelligence (AI) has sparked a revolution in the pharmaceutical industry, reshaping the way we approach complex biological challenges. “From personalized medicine to unraveling the mysteries of our DNA, we’re witnessing a new era—one where algorithms collaborate with scientists, pushing the boundaries of what’s possible,” says Seema Sayani, Ph.D., Senior Director of Life Sciences with Cognizant. “It’s accelerating drug research and development, potentially saving lives.”

Other advancements include protein structure prediction using deep learning, designing accurate 3D protein structures that aid drug design and using advanced AI techniques to generate novel molecular structures. As it continues to evolve, AI’s impact on life sciences will only grow, impacting drug discovery by predicting drug-disease associations, repurposing drugs and more. It could also help address current challenges such as false target validation, inaccurate modeling technologies and the unsuccessful communication and explanation of lab results to patients. 

“Gen AI can help address many challenges, while also optimizing processes and reducing scientists’ administrative burdens,” says Thiyagarajan Chenga Kalvinathan, Director of Drug Discovery CoE with Cognizant. Kalvinathan and Sayani highlighted four opportunities for using AI in life sciences. 

Drug discovery

The drug discovery process has evolved from manual methods to advanced automated techniques. Modern computational strategies, including target structure prediction, binding site analysis, Q-SAR predictions and virtual screening, have significantly accelerated processes. Notably, AI assists in predicting physicochemical and ADMET properties of potential drugs, thereby reducing the failure rate related to toxicity and safety issues. In as little as 40 days, researchers can now identify, synthesize, and test 20 candidates with a 10% success rate using AI, compared to 1% in the past, says Sayani.

Protein Language Models (PLM)

PLMs analyze the intricate code of life inscribed within protein sequences just as natural language models (NLP) process human language. They excel at tasks such as de novo protein sequence generation, controllable protein design, and protein property prediction. Like natural language models, protein language models are trained on vast datasets. “The datasets are based on protein sequences that contain the instructions for creating proteins,” says Sayani. “PLM learn from these sequences and gain insights into the underlying syntax of proteins to predict how different variations impact their function, stability and interactions.” 

PLMs can analyze protein sequences to identify potential drug targets, accelerating the drug discovery process and predicting protein-ligand interactions. This can enable both the virtual screening of compound libraries for drug candidates and the assessment of potential side effects of drug candidates based on their protein interactions.

RNA folding in drug discovery

AI-based methods like DRfold have improved the accuracy of RNA models by more than 70%. This can aid in the design of RNA-targeted small molecules. 

Sparked by the commercial success of RNA vaccines for COVID-19, biotechnology firms are now pursuing therapeutics based on engineered circular RNA (circRNA). RNA, in its usual linear form, is short-lived. Enzymes in cells can degrade it within hours. circRNA’s increased stability could enhance its therapeutic potential, even at low dose levels. “RNA molecules can adopt specific 3D motifs that are now considered druggable,” says Sayani. “These offer untapped potential to therapeutically modulate numerous cellular processes, including those linked to ‘undruggable’ protein targets.” 

While we’re still in early days, there’s hope that circRNA could one day lead to several new products, from next-generation vaccines and rare-disease treatments to anti-cancer agents and beyond. 

Literature search to knowledge management hub 

The influx of information during the literature review portion of the drug discovery research process can overwhelm researchers. AI-based tools can help by leveraging technologies such as text mining and natural language processing to automate key pieces. 

“Generative AI helps reduce cognitive load by summarizing, categorizing and highlighting key points and then organizing and structuring that information,” says Thiyagarajan. “We see the literature review timeline reduced by 40% to 60% using this approach.”

For example, “knowledge graphs” (visual representations of relationships between drugs, proteins, diseases and other entities) can help by contextualizing data, aiding in hypothesis generation and decision-making, as well as predicting relationships between entities and drug repurposing during target prioritization.

A new frontier

Historically, drug development followed a linear path. Researchers would identify a target, screen existing compounds and optimize them to create new drugs. With AI and other new breakthroughs, we now have the capacity to synthesize entirely novel molecules and do so in an accelerated manner. 

When Sayani thinks about the future, she says it feels boundless. “There’s a lot left to learn,” Sayani says, “but recent AI innovations have transformed the way we think about what’s possible in science and medicine.”

To learn more about Cognizant’s Life Sciences practice, please visit: cognizant.com/lifesciences or reach out to Dr. Seema Sayani, Ph.D. at [email protected] or Thiyagarajan Chenga Kalvinathan at [email protected].