The cost of discovering a new medicine increased drastically from $1.8 billion to $2.5 billion. A popular drug discovery uses virtual screening in which computer programs have a role to play. These programs need to find a compound that will react in the desired way. But there is a challenge when it comes to analyzing structures in drug design; this is where Google’s new machine learning-based initiative came into frame.
Machine learning analyses the structure of drug design and typically involves a training set to identify the known active and known inactive compounds. Next is to find the probability that a compound is active and ranking each compound based on the probability of being active. Google joined hands with pharma partners so that the machine learning can accelerate the drug delivery process. They demonstrated a virtual screening that uses Graph Convolutional Neural Networks.
To demonstrate an effective new method, Google’s accelerated science team teamed with X –Chem Pharmaceuticals. To find the active molecules, the combination of physical screening with DNA-encoded small-molecule libraries and virtual screening with the application of graph convolutional neural networks is used.
Chemome initiative, a cooperative project between Google and ZebiAI to enable smaller molecule chemical probes for biological research. A chemical library consists of stored chemicals with essential information.
A graph convolutional neural network predicts whether an arbitrarily small molecule will interact with a certain protein. The chemome initiative efficiently delivers new chemical probes to the research community.
The future direction
The biopharmaceutical industry is facing extraordinary challenges to its fundamental business models and innovation should be sufficient enough to sustain in the current setting. Research & Development must enhance their productivity and overcome the challenges in the industry. R&D investments both financial and intellectual must embrace the transformation from a traditional biopharmaceutical towards contemporary initiatives. As mentioned earlier, the Chemome initiative will spur significant new biological innovations and accelerate new therapeutic discovery for the world.
With appropriate improvements and innovations in Machine learning will enable a plethora of discoveries in pharmaceutical industries. It unlocks new opportunities and creative initiatives for a better future.
