The database of 200 million protein-structure predictions now includes homodimers, adding new biological relevance.
New paired studies from the University of Minnesota Twin Cities show that machine learning can improve the prediction of floods. The studies, published in Water Resources Research and the Proceedings ...