BioTech Startup Helps to Rescue Failed Cancer Drugs Using Neural Networks
Tue, April 20, 2021

BioTech Startup Helps to Rescue Failed Cancer Drugs Using Neural Networks

Researchers are considering drug re-use to save millions of dollars spent on a drug that failed to win approval / Credits: Gorodenkoff via Shutterstock

 

The healthcare industry is counting on AI to speed up drug discovery and development which usually takes years to complete. Hundreds of millions are being spent on every drug that researchers are working on to cure a disease or illness. However, many drugs fail to win approval, so starting from scratch is a big deal. Thus, some scientists thought that drug re-use should be considered. After all, some drugs have passed safety screening.

Lantern Pharma, a five-year-old, privately held biotech startup, hopes to save all the development money on drug development and streamline its own outlay using machine learning and neural networks. While they can’t run massive, multimillion-dollar drug trials on their own, they can test drugs in simulation before a trial happens and then partner with larger firms. The startup wants to take existing drugs and secure new use for them as more finely tailored cancer-fighting agents. 

According to ZDNet, a business technology news website that covers breaking news, analysis, and research keeps business technology professionals in touch with the latest IT trends, issues and events, Lantern will use the "NeuralNet" package to build a neural network of multiple layers and train it to predict drug response based on the presence of the genes picked out in step one. Panna Sharma, the chief executive of Lantern, and her team compared the neural network with other machine learning approaches such as "random forests," "support vector machine," and "K nearest neighbors.” 

The researchers reported that the neural network "provides faster and more accurate predictions than others.” "From our research and experiments, it is evident that ANN [artificial neural network] is the most optimal algorithm capable of capturing non-linear relationships as well as providing higher prediction accuracy even for small or limited datasets when compared to other models,” Sharma said. 

The neural network helps in creating the train, validation, and test data sets; using "grid search" to tune the hyper-parameters of the neural net; performing "stratified three-fold cross-validation" (to avoid over-fitting), and more that can help in drug-reusing. This shows that the neural network can be a valuable tool for drug discovery and development.