UCLA researchers in the Department of Chemistry and Biochemistry have developed a novel method for an artificial neural network to solve DNA problems of regression and classification.
BACKGROUND:
Organisms have diverse cell types that need to be distinguished in a clinical setting. Different cell types are known to express different molecular markers that have distinct physiological functions. Therefore, a robust method to quickly and accurately identify and quantify large quantities of cell types is important. Current methods are limited in their accuracy and limitations in fluorophores for imaging remains. A technology with a built-in framework that can identify cell-type specific DNA sequences and therefore reduce the amount of fluorophores required can increase the precision of cell type identity.
INNOVATION:
Researchers at UCLA led by Dr. Roy Wollman have developed a multilayer perceptron (MLP), a specific type of artificial neural network, which uses DNA nucleic acids that can be used to solve specific problems of biological classification. Specifically, this method can be used for cell-type mapping in conjunction with single-cell transcriptomics with a supervised machine learning algorithm to spatially map cell types. This method allows for optimal design of routine laboratory techniques such as FISH (fluorescence in situ hybridization), which can optimize the design and amount of DNA based probes to robustly identify cell types.
POTENTIAL APPLICATIONS:
ADVANTAGES:
DEVELOPMENT-TO-DATE:
This technology has been submitted for a U.S. patent.