Protein Structure Prediction Using a Maximum Likelihood Formulation of a Recurrent Geometric Network
G. Qi, M. R. Tollefson, R. A. Gogal, R. J. H. Smith, M. AlQuraishi, M. J. Schnieders
This paper improves a protein structure prediction model by incorporating experimental uncertainty data and a maximum likelihood loss function, resulting in more accurate predictions and better convergence.
bioRxiv · Sep 3, 2021 · PreprintBiophysics