14-Oct-2021 | Market Research Store
Drug manufacturing, as we all know, is a time-consuming process. The estimated count of compounds to be employed as initial points for drug development is unmanageable, which is around 1060–10100. Identifying a specific therapeutic molecule with the desired characteristics from the bulk is near to impossible.
One still has to identify and design the molecule in a laboratory and evaluate its purpose in progressively complex research trials. This is not only a laborious process but also demands hefty funds.
De novo synthesis of molecules offers the additional benefit of formulating novel chemicals that are particularly suited to specific objectives, maybe including those that no one else has explored so far. The conventional approach of designing biologically active compounds involves a specific set of regulations. The atomic or molecular fragments can be assembled to form a novel compound that is not only stable but also holds the desired biological characteristics.
The scientists used deep learning models derived from (natural language processing) NLP along with Artificial Intelligence to virtually represent a chemical molecule. The alignment of the chemicals has to be characterized as a sequence of words before incorporating NLP in drug designing. Francesca Grisoni, a researcher in the field of automating drug designing, mentioned that the piled datasets for deep learning under drug discovery are rare. In other cases, one is aware of just a few chemicals that are proven to work with a particular target.
Initially, the scientists have merged as a “rule-free” approach of deep learning to synthesize biological compounds usingon-chip, which is a type of compact automated synthesis that lowers the cost of manual work. The scientists employed a technique called transfer learning to evade the issue of limited data.
Eventually, Grisoni and her coworkers discovered 12 new bioactive molecules for liver X receptors, which have evolved as interesting targets for therapeutic compounds due to their function in inflammation regulation and lipid metabolism.
You may also like: