Search

Faithful genome partition during cell division relies on proper congression of chromosomes to the spindle equator before sister chromatid segregation. Here we uncover a kinesin-7 motor, kinetochore-associated kinesin 1 (KAK1), that is required for mitotic chromosome congression in Arabidopsis. KAK1 associates dynamically with kinetochores throughout mitosis. Loss of KAK1 results in severe defects in chromosome congression at metaphase, yet segregation errors at anaphase are rarely observed. KAK1 specifically interacts with the spindle assembly checkpoint protein BUB3.3 and both proteins show interdependent kinetochore localization. Chromosome misalignment in BUB3.3-depleted plants can be rescued by artificial tethering of KAK1 to kinetochores but not vice versa, demonstrating that KAK1 acts downstream of BUB3.3 to orchestrate microtubule-based chromosome transport at kinetochores. Moreover, we show that KAK1’s motor activity is essential for driving chromosome congression to the metaphase plate. Thus, our findings reveal that plants have repurposed BUB3.3 to interface with a specialized kinesin adapted to integrate proper chromosome congression and checkpoint control through a distinct kinetochore design.

Due to the wide applications, spam detection has long been a hot research topic in both academia and industry. Existing studies show that behavior features are effective in distinguishing the spam and legitimate reviews. However, it usually takes a long time to collect such features and thus is hard to apply them to cold-start spam review detection tasks. Recent advances leveraged the neural network to encode the various types of textual, behavior, and attribute information for this task. However, the inherent problem, i.e., lack of effective behavior features for new users who post just one review, is still unsolved. In this paper, we exploit the generative adversarial network (GAN) for addressing this problem. The key idea is to generate synthetic behavior features (SBFs) for new users from their easily accessible features (EAFs). Specifically, we first select six well recognized real behavior features (RBFs) existing for regular users. We then train a GAN framework including a generator to generate SBFs from their EAFs including text, rating, and attribute features, and a discriminator to discriminate RBFs and SBFs. We design a new implementation of generator and discriminator for effective training. The trained GAN is finally applied to new users for generating synthetic behavior features. We conduct extensive experiments on two Yelp datasets. Experimental results demonstrate that our proposed framework significantly outperforms the state-of-the-art methods. [ABSTRACT FROM AUTHOR]

In the field of biological science, correct field identification of species is of critical importance and requires effective access to and use of specialized biological information. Field guides, floras and faunas, and other publications are written to provide critical information for identification. However, fulltext retrieval systems are ineffective at providing access to the very specific information needed for species identification. In this study, dichotomous keylike information that is often used in species identification is automatically generated from botanical documents and used to enhance fulltext botanical document retrieval. Such information is used to describe document content and is also used to provide a formbased search that allows users to apply their domain knowledge to define various plant characteristics in their queries. Information extraction techniques were used to generate the keylike information from semistructured botanical documents. Experimental results indicate that this approach improved search performance as well as users satisfaction.

This study analyzed the user search queries of an experiment conducted on a fulltext botanical retrieval system. Search terms were extracted and categorized. Term frequencies in each category were calculated. Search terms were compared with target documents. The results reveal preferences for the kinds of information used in search queries, that search queries and document representations use very different format and vocabulary, and that the domain knowledge used in user search queries is lost when such queries are interpreted by the system.