ABSTRACT
An amendment to this paper has been published and can be accessed via a link at the top of the paper.
ABSTRACT
An amendment to this paper has been published and can be accessed via a link at the top of the paper.
ABSTRACT
An amendment to this paper has been published and can be accessed via a link at the top of the paper.
ABSTRACT
Pinpointing subcellular protein localizations from microscopy images is easy to the trained eye, but challenging to automate. Based on the Human Protein Atlas image collection, we held a competition to identify deep learning solutions to solve this task. Challenges included training on highly imbalanced classes and predicting multiple labels per image. Over 3 months, 2,172 teams participated. Despite convergence on popular networks and training techniques, there was considerable variety among the solutions. Participants applied strategies for modifying neural networks and loss functions, augmenting data and using pretrained networks. The winning models far outperformed our previous effort at multi-label classification of protein localization patterns by ~20%. These models can be used as classifiers to annotate new images, feature extractors to measure pattern similarity or pretrained networks for a wide range of biological applications.
Subject(s)
Deep Learning , Image Processing, Computer-Assisted/methods , Microscopy, Fluorescence/methods , Proteins/analysis , HumansABSTRACT
UNLABELLED: Tropical is a software for simulation and parameter estimation of reaction-diffusion models. Based on spatio-temporal microscopy images, Tropical estimates reaction and diffusion coefficients for user-defined models. Tropical allows the investigation of systems with an inhomogeneous distribution of molecules, making it well suited for quantitative analyses of microscopy experiments such as fluorescence recovery after photobleaching (FRAP). AVAILABILITY: Tropical is available free of charge for academic use at http://www.dkfz.de/tbi/projects/modellingAndSimulationOfCelluarSystems/tropical.jsp after signing a material transfer agreement.
Subject(s)
Cell Physiological Phenomena , Image Interpretation, Computer-Assisted/methods , Microscopy, Fluorescence/methods , Models, Biological , Protein Interaction Mapping/methods , Signal Transduction/physiology , Software , Algorithms , Computer Simulation , DiffusionABSTRACT
Promyelocytic leukemia nuclear bodies (PML NBs) have been proposed to be involved in tumor suppression, viral defense, DNA repair, and/or transcriptional regulation. To study the dynamics of PML NBs during mitosis, we developed several U2OS cell lines stably coexpressing PML-enhanced cyan fluorescent protein with other individual marker proteins. Using three-dimensional time-lapse live cell imaging and four-dimensional particle tracking, we quantitatively demonstrated that PML NBs exhibit a high percentage of directed movement when cells progressed from prophase to prometaphase. The timing of this increased dynamic movement occurred just before or upon nuclear entry of cyclin B1, but before nuclear envelope breakdown. Our data suggest that entry into prophase leads to a loss of tethering between regions of chromatin and PML NBs, resulting in their increased dynamics. On exit from mitosis, Sp100 and Fas death domain-associated protein (Daxx) entered the daughter nuclei after a functional nuclear membrane was reformed. However, the recruitment of these proteins to PML NBs was delayed and correlated with the timing of de novo PML NB formation. Together, these results provide insight into the dynamic changes associated with PML NBs during mitosis.