New method screens natural compounds

UCSD Jacobs School of Engineering researchers have invented computational tools to decode and determine whether natural compounds collected in oceans and forests are new or have already been described and are, therefore, not patentable. In particular, the advance enables the rapid characterization of ring-shaped nonribosomal peptides (NRPs), a class of pharmaceutically important natural compounds.

NRPs often serve as chemical defenses for the bacteria that manufacture them and have an unparalleled track record in pharmacology dating back to the discovery of penicillin.

Currently, however, it is difficult, time-consuming and costly to determine the molecular structure of NRPs. The new UCSD algorithms quickly make sense of the flood of tiny peptide fragments generated by mass spectrometers that blast nonribosomal peptides apart and determine their sizes. The study appears online in the journal Nature Methods.

Newly discovered protein’s role in immunity

Much is known of the cascade of signals required for immune-cell development. Now, researchers at The Scripps Research Institute (TSRI) report the discovery of a critical signal in T-cell development that has remained elusive.

Researchers made the discovery while looking for proteins produced in the thymus and dubbed the new protein Themis. To gain insights into its function, mice were engineered that lacked the corresponding gene. The result: Themis-deficient mice had fewer mature T cells compared with normal mice. The finding appears in Nature Immunology.

The new science of learning

Of all the qualities that distinguish humans from other species, how we learn is one of the most significant. From collaborative efforts in neuroscience, psychology, education and machine learning has emerged a new “science of learning” that is reshaping the concept of learning.

Research at the Temporal Dynamics of Learning Center (TDLC) at UCSD and the University of Washington points to three principles guiding human learning:

  • Learning is computational (infants and children possess powerful computational skills that allow them to infer structured models of their environment);
  • Learning is social (supported by child-robot interaction studies that show learning depends on how social and responsive the robot is); and
  • Learning is supported by brain circuits linking perception and action (borne out by discoveries in the brain systems that govern perception and production of actions).

The findings appear in the journal Science.