PRESENT DEEP LEARNING TECHNOLOGY
Deep learning is a form of machine learning that uses neural networks, layers of artificial neurons. Over large volumes of labeled data, the strength of the weights connecting the neurons is fine-tuned through gradient descent to identify patterns in the input samples correlated with particular outcomes so that the model learns to predict the correct sample outputs.
WHAT ARE THE LIMITATIONS?
Since there thousands of lysin species with many still undiscovered, it is infeasible to store them all at local centers for patients.
​
Instead of classification, deep learning will be used in LyseDevice for the on-demand generation of amino acid sequences of novel lysin therapies to treat any given infection.
DEEP LEARNING TO IMPROVE LYSIN THERAPY
Since lysins are highly specialized, time-consuming experimental testing is traditionally required to develop treatments for each bacterial species.
​
HOW?
From the DNA of the bacteria and the amino acid sequence of the lysin, recurrent neural networks are trained to classify whether the lysin can successfully target the bacteria. This is used to match a bacteria with its corresponding lysin, allowing for infection treatments to be efficiently identified from a database of lysins.
​
​
​
​
​
​
​
​