13197 research outputs found
Sort by
リッタイ ケイジョウ ノ イジ ト ゲンシ チツジョ ヒョウメン オ リョウリツシタ サンジゲン Siライン コウゾウ ノ サクセイ
Get PDF奈良先端科学技術大学院大学修士(工学)master thesi
Disentangling Pretrained Representation to Incorporate Unseen Low-Resource Languages in Multilingual Machine Translation
Get PDF奈良先端科学技術大学院大学修士(工学)master thesi
An Empirical Study on Removing Library’s Dependencies for Open-Source Libraries
Get PDF奈良先端科学技術大学院大学修士(工学)master thesi
Assaying Effector Cell-to-Cell Mobility in Plant Tissues Identifies Hypermobility and Indirect Manipulation of Plasmodesmata
No full textIn plants, plasmodesmata establish cytoplasmic continuity between cells to allow for communication and resource exchange across the cell wall. While plant pathogens use plasmodesmata as a pathway for both molecular and physical invasion, the benefits of molecular invasion (cell-to-cell movement of pathogen effectors) are poorly understood. To establish a methodology for identification and characterization of the cell-to-cell mobility of effectors, we performed a quantitative live imaging-based screen of candidate effectors of the fungal pathogen Colletotrichum higginsianum. We predicted C. higginsianum effectors by their expression profiles, the presence of a secretion signal, and their predicted and in planta localization when fused to green fluorescent protein. We assayed for cell-to-cell mobility of nucleocytosolic effectors and identified 14 that are cell-to-cell mobile. We identified that three of these effectors are “hypermobile,” showing cell-to-cell mobility greater than expected for a protein of that size. To explore the mechanism of hypermobility, we chose two hypermobile effectors and measured their impact on plasmodesmata function and found that even though they show no direct association with plasmodesmata, each increases the transport capacity of plasmodesmata. Thus, our methods for quantitative analysis of cell-to-cell mobility of candidate microbe-derived effectors, or any suite of host proteins, can identify cell-to-cell hypermobility and offer greater understanding of how proteins affect plasmodesmal function and intercellular connectivity.journal articl
Fast gathering despite a linear number of weakly Byzantine agents
Get PDFIn this work, we study the gathering problem to make multiple agents, who are initially scattered in arbitrary networks, meet at the same node. The network has k agents with unique identifiers (IDs), and f of them are weakly Byzantine agents that behave arbitrarily, except for falsifying their identifiers. These agents behave in synchronous rounds, and they may start an algorithm at different rounds. Each agent cannot leave information at a node. We propose herein a deterministic algorithm that efficiently achieves gathering with a simultaneous termination having a small number of non-Byzantine agents. The proposed algorithm concretely works in O(f00B7X(N)) rounds if the agents know the upper bound N on the number of nodes, and at least 7f+7 non-Byzantine agents exist, where |Λall| is the length of the largest ID among agents, and X(n) is the number of rounds required to explore any network composed of n nodes. The literature presents two efficient gathering algorithms with a simultaneous termination. The first algorithm assumes that agents know the number n of nodes and achieves the gathering in O(n400B7X(n)) rounds in the presence of any number of Byzantine agents, where |Λgood| is the length of the largest ID among non-Byzantine agents. The second algorithm assumes both that agents know N and that at least 4f2+8f+4 non-Byzantine agents exist, and it achieves the gathering in O((f+|Λall|)$00B7X(N)) rounds. The proposed algorithm is faster than the first existing algorithm and requires fewer non-Byzantine agents than the second existing algorithm if n is given to agents. We propose herein a new technique to simulate a Byzantine consensus algorithm for synchronous message-passing systems on agent systems to reduce the number of agents.journal articl
Assessing domain adaptation in adverse drug event extraction on real-world breast cancer records
No full textBackground: Adverse Drug Events (ADE) are key information present in unstructured portions of Electronic Health Records. These pose a significant challenge in healthcare, ranging from mild discomfort to severe complications, and can impact patient safety and treatment outcomes.
Methods: We explore the influence of domain shift between a set of dummy clinical notes and a real-world hospital corpus of Japanese clinical notes of breast cancer treatment when extracting ADEs from free text. We annotated a subset of the hospital dataset and used it to fine-tune a Named Entity Recognition (NER) model, initially trained with the set of dummy documents. We used increasing amounts of the annotated data and evaluated the impact on the model's performance. Additionally, we examined the extracted information to identify combinations of drugs that are likely to cause ADEs.
Results: We show that domain adaptation can significantly improve model performance in the new domain, as by feeding a small subset of 100 documents for the fine-tuning process we saw a 40% improvement in model performance. However, we also noticed diminishing returns when fine-tuning the model with a larger dataset. For instance, by feeding eight times more data, we only saw further 18% improvement in extraction performance.
Conclusion: While variations in writing style and vocabulary in clinical corpora can significantly impact the quality of NER results. We show that domain adaptation can be of great aid in mitigating these discrepancies and achieving better performance. Yet, while providing in-domain data to a model helps, there are diminishing returns when fine-tuning with large amounts of data.journal articl
セイソウ コウジョウナイ ニ オケル カンキョウ オン オ フィンガー プリント ト シテ モチイタ オクナイ エリア スイテイ
Get PDF奈良先端科学技術大学院大学修士(工学)master thesi
Induction of Endoplasmic Reticulum Stress by Prodigiosin in Yeast Saccharomyces cerevisiae
No full textProdigiosin, a red pigment produced by numerous bacterial species, exerts various antibiotic effects on prokaryotic and eukaryotic organisms. For instance, human carcinoma cell lines appear to suffer from endoplasmic reticulum (ER) stress in the presence of prodigiosin. Here, we demonstrated that prodigiosin also triggers the unfolded-protein response (UPR), which is a cytoprotective response against ER stress, in yeast Saccharomyces cerevisiae. An S. cerevisiae mutant carrying a UPR-deficient mutation was hypersensitive to prodigiosin. Our observations cumulatively indicate that protein folding in the ER is impaired by prodigiosin, illustrating a new mode of action.journal articl