A new versatile, autonomous, robotic experimental platform (MAPLE) can increase the throughput of biological experiments by automating the growth and phenotyping of a variety of model organisms.
Acquisition of behavioral sequences in normally aged mice involves short and unusually fast patterns of action, some of which are reproduced by striatal circuitry manipulations in young mice and can be transitorily restored through action-related feedback.
Development of a fully automated pain scale using machine learning tools in computational neuroethology and creation of new software, reveals a robust circuit-dissection compatible platform for objective pain measurement.
Automated liquid handling, whole mount staining, and clearing allow unbiased 3D quantitation of cell markers in human neural organoids with diameters of up to 1 mm at the single-cell level.
An algorithm for analysing brain connectivity data identifies cell types and connections in simple (C. elegans) and complex (mouse) nervous systems, and can even resolve structure and connectivity in a man-made microprocessor.
In an investigation into the effects of drugs on proteins, an active machine learning algorithm chose which sets of experiments to perform and was able to learn an accurate model of the effects after doing only a fraction of the experiments.
A robot capable of automatically obtaining blind whole cell patch clamp recordings from multiple neurons simultaneously guides four interacting electrodes in a coordinated fashion, avoiding mechanical coupling in the brain.
A comparison of different bioimage analysis pipelines reveals how deep learning can be used for automatized and reliable analysis of fluorescent features in biological datasets.
An open-source python package for phenotype analyses provides a versatile, modular and user-friendly solution to determine complementary fitness-related traits from large-scale assays of microbial colonies.