Identifying Proteins Faster and Cheaper
Using machine learning, nanopores, and electrodes for proteomics
Identifying proteins can be an expensive pain. Traditionally, researchers have used mass-spectrometry based technologies to figure out which one of the almost infinite number of proteins they are working with. Stuart Lindsay, a biophysicist from Arizona State University, has described a new technique that uses scanning tunneling microscopy (STM) instead.
The STM method was previously used to identify DNA base sequences. By sending DNA through a nanopore equipped with electrodes, the resultant electrical spike could be used to identify which of the four types of DNA bases just traveled through the pore. By introducing machine learning, Lindsay and his team developed a way to instead make more subtle identifications between the 20 amino acids that make up all proteins.
Now the technique can be used to identify a protein by reading each of its amino acids, just like it has been used to identify a strand of DNA by reading each of its bases. For more on this story, head over to CNN.