Among land plants, genome sizes vary remarkably by > 2,200-fold. This variation depends on the loss and gain of non-coding DNA sequences, forming different heterochromatin complexes during interphase. In plants with a giant genome, the major part of chromatin stays condensed in interphase, forming a dense meshwork of heterochromatin threads, called interphase chromonemata. Using super-resolution light and electron microscopy, we studied the ultrastructure of chromonemata during and after replication in root meristem nuclei of Nigella damascena L. During S-phase, heterochromatin undergoes transient decompaction locally at the active sites of DNA synthesis, and due to the heterochromatin abundance, chromonema replication is accompanied by a robust chromonema meshwork disassembly, which led to the general reorganization of the nucleus morphology visible even by conventional light microscopy. After replication, the heterochromatin condenses again, restoring the chromonema structure. Thus, we showed that heterochromatin replication in interphase nuclei of giant-genome plants induces a global chromonema decondensation and reorganization.

No abstract

The article focuses on consideration of complex aerospace monitoring of aggregated territories. It is made to assess the processes took place in "lithosphere-atmosphere-ionosphere-magnetosphere" system when exploration of oil and gas fields. The research is based on the remote aerospace sounding's method. As a result of this research we have an access to the complex analysis of huge range of data. We have an opportunity to calculate a "convolution operation" in the systems of complex aerospace monitoring in order to identify the signs that can help to indicate the territory having oil and gas deposits. To sum up, the research reveals the correlation between the huge range of anomalies of the nature and the presence of oil and gas deposits in the earth on the basis of remote aerospace monitoring sounding method.

Animals have remarkable abilities to adapt locomotion to different terrains and tasks. However, robots trained by means of reinforcement learning are typically able to solve only a single task and a transferred policy is usually inferior to that trained from scratch. In this work, we demonstrate that meta-reinforcement learning can be used to successfully train a robot capable to solve a wide range of locomotion tasks. The performance of the meta-trained robot is similar to that of a robot that is trained on a single task.

No abstract