The database of 200 million protein-structure predictions now includes homodimers, adding new biological relevance.
Machine learning analysis reveals which metrics drive March Madness seeding and predictive analytics in committee decisions.