It is extremely exciting to see artificial intelligence and other techniques being used to support drug discovery.
FREMONT, CA: According to reports, 90 percent of all drug possibilities fail. The few that succeed take an average of 10 years to reach the market and cost from $2.5 billion to $12 billion to get there. The majority of this overall cost and time is attributed to the drug discovery phase needing the synthetization of thousands of molecules to generate a single pre-clinical lead candidate. Hence, researchers are searching for ways to leverage artificial intelligence-based models to expedite this discovery phase at a significantly lower cost. Read on to know more.
Deep generative models, like variational autoencoders and generative adversarial networks, are considered promising for the computational creation of novel molecules due to their results in the virtual synthesis of images, text, speech, and image captions. Virtual creation of new and optimal lead candidates needs exploring and performing a multi-objective optimization in large chemical space, as the model requires assessment and balance between critical factors like drug activity, stability, selectivity, toxicity, ease of synthesis, etc.
Such multi-objective optimization is managed using either conditional generative models or optimization ways like Bayesian optimization.
Generating new and optimal antimicrobial peptides by learning from a limited repository of known AMP sequences is a challenging task. This research is critical, given that AMPs are viewed as a drug of last resort against antimicrobial resistance, one of the major threats to global health, food security, and development. It is thought that bacterial co-infections and widespread antibiotic use could further antibiotic fuel resistance worldwide.
It is exciting to see deep learning and other techniques being used to pinpoint drug discovery in a matter of days. In particular, exploiting large, publicly-available data to accelerate this process can give huge benefits for a reduced cost. The data-driven method will give better and faster results than the legacy methods, leading to faster drug discovery and safer, more reliable outcomes than clinical trials on their own. While it's unlikely that AI will replace the present methods overnight, it's obvious that organizations that add AI to their methods will rapidly replace those who do not.