Explore the vast range of titles available.

Manga (Japanese comics) are popular worldwide. However, current e-manga archives offer very limited search support, i.e., keyword-based search by title or author. To make the manga search experience more intuitive, efficient, and enjoyable, we propose a manga-specific image retrieval system. The proposed system consists of efficient margin labeling, edge orientation histogram feature description with screen tone removal, and approximate nearest-neighbor search using product quantization. For querying, the system provides a sketch-based interface. Based on the interface, two interactive reranking schemes are presented: relevance feedback and query retouch. For evaluation, we built a novel dataset of manga images, Manga109, which consists of 109 comic books of 21,142 pages drawn by professional manga artists. To the best of our knowledge, Manga109 is currently the biggest dataset of manga images available for research. Experimental results showed that the proposed framework is efficient and scalable (70 ms from 21,142 pages using a single computer with 204 MB RAM).

Manga (Japanese comics) are popular worldwide. However, current e-manga archives offer very limited search support, including keyword-based search by title or author, or tag-based categorization. To make the manga search experience more intuitive, efficient, and enjoyable, we propose a content-based manga retrieval system. First, we propose a manga-specific image-describing framework. It consists of efficient margin labeling, edge orientation histogram feature description, and approximate nearest-neighbor search using product quantization. Second, we propose a sketch-based interface as a natural way to interact with manga content. The interface provides sketch-based querying, relevance feedback, and query retouch. For evaluation, we built a novel dataset of manga images, Manga109, which consists of 109 comic books of 21,142 pages drawn by professional manga artists. To the best of our knowledge, Manga109 is currently the biggest dataset of manga images available for research. We conducted a comparative study, a localization evaluation, and a large-scale qualitative study. From the experiments, we verified that: (1) the retrieval accuracy of the proposed method is higher than those of previous methods; (2) the proposed method can localize an object instance with reasonable runtime and accuracy; and (3) sketch querying is useful for manga search.

BackgroundSurgical resection is the only potentially curative therapy for gallbladder carcinoma (GBC). However, the postoperative recurrence rate is high (approximately 50%), and recurrence occasionally develops early after surgery.Patients and MethodsA total of 139 patients who underwent macroscopically curative resection for GBC between 2002 and 2018 were retrospectively reviewed. Early recurrence (ER) was defined as recurrence within 6 months after surgery. Univariate and multivariate logistic regression analysis was performed using preoperative factors that may influence early recurrence, namely patient background factors, tumor markers, imaging findings, and body composition parameters obtained preoperatively, to create a predictive score for ER.ResultsThe median follow-up period was 21.9 months (range, 6.2–195.7 months). Postoperative recurrence was observed in 55 (39.6%) patients, of whom 14 (25.5%) developed ER. The median overall survival after surgery was 104.7 months for the non-ER group and 15.7 months for the ER group. On multivariate analysis, high carbohydrate antigen 19-9, low muscle attenuation, high visceral fat attenuation, liver invasion, and other organ invasion on preoperative computed tomography were identified as independent risk factors for ER. A preoperatively predictive scoring system for ER was constructed by weighting the above five factors. The nomogram showed an area under the curve of 0.881, indicating good predictive potential for ER.ConclusionsER in resected GBC indicates a very poor prognosis. The present preoperative scoring system can sufficiently predict ER and may be helpful in determining the optimal treatment strategies.

Plant–microbe interactions strongly influence plant growth and nutrient acquisition, and their outcomes depend on nutrient availability. The root endophyte Colletotrichum tofieldiae ( Ct ) promotes growth in Arabidopsis thaliana under inorganic phosphate (Pi) limitation, but its effects under Pi sufficiency and the role of salicylic acid (SA) signaling remain unclear. Here, we examined Pi-dependent growth responses, nutrient accumulation, and SA signaling in wild-type (WT) and SA-deficient ics1 mutant plants co-cultivated with Ct under low, moderate, and high Pi conditions (25, 150, and 625 µM). Under low Pi, Ct significantly enhanced WT growth, increasing leaf number and root length by 41.8% and 50.5%, respectively, and promoting biomass accumulation, with fresh and dry weight increases of 104% and 232% relative to uninoculated controls. Growth promotion was reduced at moderate Pi and shifted toward growth suppression under high Pi. Elemental profiling using inductively coupled plasma mass spectrometry (ICP-MS) revealed pronounced Ct -mediated nutrient accumulation under Pi limitation. At low Pi, phosphorus content increased by 281%, accompanied by significant increases in K (70.1%), S (84.5%), and Ca (73.2%). In contrast, at moderate and high Pi, Ct consistently enhanced P accumulation, while changes in K, S, and Ca were not significant. Ct colonization induced expression of the SA-responsive marker gene PR1 , particularly under low Pi. In contrast, ics1 mutants failed to exhibit Ct- induced growth promotion and instead displayed growth suppression across all Pi conditions. Together, these findings demonstrate that Pi availability and ICS1-mediated SA biosynthesis jointly determine the outcome of the Arabidopsis – Ct interaction.

The β -chemokine receptor CCR5 is considered to be an attractive target for inhibition of macrophage-tropic (CCR5-using or R5) HIV-1 replication because individuals having a nonfunctional receptor (a homozygous 32-bp deletion in the CCR5 coding region) are apparently normal but resistant to infection with R5 HIV-1. In this study, we found that TAK-779, a nonpeptide compound with a small molecular weight (Mr 531.13), antagonized the binding of RANTES (regulated on activation, normal T cell expressed and secreted) to CCR5-expressing Chinese hamster ovary cells and blocked CCR5-mediated Ca2+ signaling at nanomolar concentrations. The inhibition of β -chemokine receptors by TAK-779 appeared to be specific to CCR5 because the compound antagonized CCR2b to a lesser extent but did not affect CCR1, CCR3, or CCR4. Consequently, TAK-779 displayed highly potent and selective inhibition of R5 HIV-1 replication without showing any cytotoxicity to the host cells. The compound inhibited the replication of R5 HIV-1 clinical isolates as well as a laboratory strain at a concentration of 1.6-3.7 nM in peripheral blood mononuclear cells, though it was totally inactive against T-cell line-tropic (CXCR4-using or X4) HIV-1.

BackgroundT cell immunoglobulin and mucin domain-containing protein 3 (TIM-3) is a part of modules expressed on dysfunctional or exhausted T cells as well as dendritic cells and has emerged as a target for several therapeutic antibodies that are under clinical development. Co-blockade of TIM-3 and PD-1 results in tumor regression in preclinical models and improves anticancer T cell responses in patients with advanced cancers. TIM-3 has been reported to have multiple ligands including galectin-9, phosphatidylserine, CEACAM-1 and HMGB1, which bind to different regions on the extracellular domain of TIM-3. Most of the TIM-3 antibodies developed to date are intended to inhibit phosphatidylserine that binds to the pocket in TIM-3 immunoglobulin V domain. Galectin-9 binds to carbohydrate motifs on the opposite side of phosphatidylserine-binding site in immunoglobulin V domain and thereby induces cell death in TIM-3+ T cells. We report herein novel antibodies that block TIM-3 binding to multiple ligands including these two important ligands simultaneously.MethodsAnti-TIM-3 antibodies were generated by immunizing mice with a purified recombinant TIM-3 protein and TIM-3-expressing mammalian cell line. Phage display libraries were constructed using cDNAs of splenocytes and lymph node cells of the immunized mice, then subjected to the biopanning using recombinant TIM-3 proteins. After analyzing specificities and affinities to the TIM-3 protein, scFvs obtained were classified by epitope bin and inhibitory effects on TIM-3 binding to the multiple ligands. The scFvs were converted to scFv-Fc to generate biparatopic (bispecific) antibodies.ResultsAt least five classes of TIM-3 antibodies were obtained, and each class was grouped into different epitope bins and has unique inhibitory profiles for multiple ligands of TIM-3. Their biparatopic (bispecific) forms were produced from the scFv clones and subjected to the analyses of TIM-3 binding, inhibition of ligand binding, and immune activation. As expected, the biparatopic antibodies that recognize two different epitopes showed higher affinity and specificity to TIM-3 than monospecific forms. A lead biparatopic antibody that block the binding of TIM-3 to galectin-9 and phosphatidylserine showed remarkable potency on T cell activation, protection from exhaustion and apoptotic cell death of T cells as well as more potent anti-tumor efficacy.ConclusionsThis study demonstrates the successful development of a novel biparatopic antibody that blocks the binding of TIM-3 to phosphatidylserine and galectin-9 simultaneously. The antibody shows the advantages over conventional TIM-3 antibodies in reducing T cell exhaustion and potentially manipulated for the development of human monoclonal antibodies for therapeutic treatment of cancer.

User-generated data sources have gained significance in uncovering Adverse Drug Reactions (ADRs), with an increasing number of discussions occurring in the digital world. However, the existing clinical corpora predominantly revolve around scientific articles in English. This work presents a multilingual corpus of texts concerning ADRs gathered from diverse sources, including patient fora, social media, and clinical reports in German, French, and Japanese. Our corpus contains annotations covering 12 entity types, four attribute types, and 13 relation types. It contributes to the development of real-world multilingual language models for healthcare. We provide statistics to highlight certain challenges associated with the corpus and conduct preliminary experiments resulting in strong baselines for extracting entities and relations between these entities, both within and across languages.

A computer program package named \"BRAIN\" has been developed to simulate the kinematics and the performance of rolling bearings under various running conditions. The calculation time necessary for running BRAIN software on a PC is very short. Various outputs can be obtained using BRAIN such as running torque, roller skew angle, roller slippage, and PV values. Several experiments have been conducted to confirm the validity of BRAIN. The running torque of a four point contact ball bearing and that of a tapered roller bearing were measured. In addition, the skew of the roller in a needle bearing was measured. These experimental results were compared with the calculation results. The experiments and the calculations showed good agreement.