A decoding-based, state-space reconstruction reveals that neurons in macaque IT cortex change the structure of their collective attractor dynamics depending on task contexts.
Text mining of complete EHRs for 14,017 diabetes patients and subsequent clustering led to phenotypically deep clusters, showing distinct glycemic profiles, comorbidities, and SNP association patterns.
A two-part neural network models reward-based training and provides a unified framework in which to study diverse computations that can be compared to electrophysiological recordings from behaving animals.
A biologically plausible learning rule enables recurrent neural networks to model the way in which neural circuits use supervised learning to perform time-dependent computations.
A general machine learning scheme for integrating time-series data from single-molecule experiments and molecular dynamics simulations is proposed and successfully demonstrated for the folding dynamics of the WW domain.
Metabolite analysis of plasma from enteric fever patients define signals of organism specific host–pathogen interactions and provides opportunities for new diagnostics.
Cerebellar climbing fibers can generate learned reward-predictive instructional signals, suggesting a role for cerebellar learning in the reinforcement of reward-driven behaviors.
Easy-to-use image analysis software enables single cell quantitation of cell types and division rates in complex 3D tissues including living Drosophila brains, mouse embryos and Zebrafish organoids.
