Multi-level attention pooling for graph neural networks: Unifying graph representations with multiple localities.

Graph neural networks (GNNs) have been widely used to learn vector representation of graph-structured data and achieved better task performance than conventional methods. The foundation of GNNs is the message passing procedure, which propagates the information in a node to its neighbors. Since this procedure proceeds one step per layer, the range of the information propagation among nodes is small in the lower layers, and it expands toward the higher layers. Therefore, a GNN model has to be deep enough to capture global structural information in a graph. On the other hand, it is known that deep GNN models suffer from performance degradation because they lose nodes' local information, which would be essential for good model performance, through many message passing steps. In this study, we propose multi-level attention pooling (MLAP) for graph-level classification tasks, which can adapt to both local and global structural information in a graph. It has an attention pooling layer for each message passing step and computes the final graph representation by unifying the layer-wise graph representations. The MLAP architecture allows models to utilize the structural information of graphs with multiple levels of localities because it preserves layer-wise information before losing them due to oversmoothing. Results of our experiments show that the MLAP architecture improves the graph classification performance compared to the baseline architectures. In addition, analyses on the layer-wise graph representations suggest that aggregating information from multiple levels of localities indeed has the potential to improve the discriminability of learned graph representations.

Implicit Contact Dynamics Modeling With Explicit Inertia Matrix Representation for Real-Time, Model-Based Control in Physical Environment.

Model-based control has great potential for use in real robots due to its high sampling efficiency. Nevertheless, dealing with physical contacts and generating accurate motions are inevitable for practical robot control tasks, such as precise manipulation. For a real-time, model-based approach, the difficulty of contact-rich tasks that requires precise movement lies in the fact that a model needs to accurately predict forthcoming contact events within a limited length of time rather than detect them afterward with sensors. Therefore, in this study, we investigate whether and how neural network models can learn a task-related model useful enough for model-based control, that is, a model predicting future states, including contact events. To this end, we propose a structured neural network model predictive control (SNN-MPC) method, whose neural network architecture is designed with explicit inertia matrix representation. To train the proposed network, we develop a two-stage modeling procedure for contact-rich dynamics from a limited number of samples. As a contact-rich task, we take up a trackball manipulation task using a physical 3-DoF finger robot. The results showed that the SNN-MPC outperformed MPC with a conventional fully connected network model on the manipulation task.

Development of an SSR marker-based genetic linkage map and identification of a QTL associated with flowering time in Eustoma.

Lisianthus (Eustoma grandiflorum) is an important floricultural crop cultivated worldwide. Despite its commercial importance, few DNA markers are available for molecular genetic research. In this study, we constructed a genetic linkage map and to detect quantitative trait loci (QTLs) for important agronomic traits of lisianthus. To develop simple sequence repeat (SSR) markers, we used 454-pyrosequencing technology to obtain genomic shotgun sequences and subsequently identified 8263 putative SSRs. A total of 3990 primer pairs were designed in silico and 1189 unique primer pairs were extracted through a BLAST search. Amplification was successful for more than 1000 primer pairs, and ultimately 278 SSR markers exhibited polymorphism between the two lisianthus accessions evaluated. Based on these markers, a genetic linkage map was constructed using a breeding population derived from crosses between the two accessions, for which flowering time differed (>140 days when grown under 20°C). We detected one QTL associated with flowering time (phenotypic variance, 27%; LOD value, 3.7). The SSR marker located at this QTL may account for variation in flowering time among accessions (i.e., three accessions whose nodes of the first flower were over 30 had late-flowering alleles of this QTL).

Optimal Scheduling for Laboratory Automation of Life Science Experiments with Time Constraints.

In automated laboratories consisting of multiple different types of instruments, scheduling algorithms are useful for determining the optimal allocations of instruments to minimize the time required to complete experimental procedures. However, previous studies on scheduling algorithms for laboratory automation have not emphasized the time constraints by mutual boundaries (TCMBs) among operations, which is important in procedures involving live cells or unstable biomolecules. Here, we define the "scheduling for laboratory automation in biology" (S-LAB) problem as a scheduling problem for automated laboratories in which operations with TCMBs are performed by multiple different instruments. We formulate an S-LAB problem as a mixed-integer programming (MIP) problem and propose a scheduling method using the branch-and-bound algorithm. Simulations show that our method can find the optimal schedules of S-LAB problems that minimize overall execution time while satisfying the TCMBs. Furthermore, we propose the use of our scheduling method for the simulation-based design of job definitions and laboratory configurations.

S-540956, a CpG Oligonucleotide Annealed to a Complementary Strand With an Amphiphilic Chain Unit, Acts as a Potent Cancer Vaccine Adjuvant by Targeting Draining Lymph Nodes.

Robust induction of cancer-antigen-specific CD8+ T cells is essential for the success of cancer peptide vaccines, which are composed of a peptide derived from a cancer-specific antigen and an immune-potentiating adjuvant, such as a Toll-like receptor (TLR) agonist. Efficient delivery of a vaccine antigen and an adjuvant to antigen-presenting cells in the draining lymph nodes (LNs) holds key to maximize vaccine efficacy. Here, we developed S-540956, a novel TLR9-agonistic adjuvant consisting of B-type CpG ODN2006 (also known as CpG7909), annealed to its complementary sequence oligodeoxynucleotide (ODN) conjugated to a lipid; it could target both a cancer peptide antigen and a CpG-adjuvant in the draining LNs. S-540956 accumulation in the draining LNs and activation of plasmacytoid dendritic cells (pDCs) were significantly higher than that of ODN2006. Mechanistic analysis revealed that S-540956 enhanced the induction of MHC class I peptide-specific CD8+ T cell responses via TLR9 in a CD4+ T cell-independent manner. In mice, the therapeutic effect of S-540956-adjuvanted with a human papillomavirus (HPV)-E7 peptide vaccine against HPV-E7-expressing TC-1 tumors was significantly better than that of an ODN2006-adjuvanted vaccine. Our findings demonstrate a novel adjuvant discovery with the complementary strand conjugated to a lipid, which enabled draining LN targeting and increased ODN2006 accumulation in draining LNs, thereby enhancing the adjuvant effect. Our findings imply that S-540956 is a promising adjuvant for cancer peptide vaccines and has a high potential for applications in various vaccines, including recombinant protein vaccines.

Simultaneous induction of mutant alleles of two allergenic genes in soybean by using site-directed mutagenesis.

BACKGROUND: Soybean (Glycine max) is a major protein crop, because soybean protein has an amino acid score comparable to that of beef and egg white. However, many allergens have been identified among soybean proteins. A decrease in allergenic protein levels would be useful for expanding the market for soybean proteins and processed foods. Recently, the CRISPR/Cas9 system has been adopted as a powerful tool for the site-directed mutagenesis in higher plants. This system is expected to generate hypoallergenic soybean varieties. RESULTS: We used two guide RNAs (gRNAs) and Agrobacterium-mediated transformation for simultaneous site-directed mutagenesis of two genes encoding the major allergens Gly m Bd 28 K and Gly m Bd 30 K in two Japanese soybean varieties, Enrei and Kariyutaka. We obtained two independent T0 Enrei plants and nine T0 Kariyutaka plants. Cleaved amplified polymorphic sequence (CAPS) analysis revealed that mutations were induced in both targeted loci of both soybean varieties. Sequencing analysis showed that deletions were the predominant mutation type in the targeted loci. The Cas9-free plants carrying the mutant alleles of the targeted loci with the transgenes excluded by genetic segregation were obtained in the T2 and T3 generations. Variable mutational spectra were observed in the targeted loci even in T2 and T3 progenies of the same T0 plant. Induction of multiple mutant alleles resulted in six haplotypes in the Cas9-free mutants derived from one T0 plant. Immunoblot analysis revealed that no Gly m Bd 28 K or Gly m Bd 30 K protein accumulated in the seeds of the Cas9-free plants. Whole-genome sequencing confirmed that a Cas9-free mutant had also no the other foreign DNA from the binary vector. Our results demonstrate the applicability of the CRISPR/Cas9 system for the production of hypoallergenic soybean plants. CONCLUSIONS: Simultaneous site-directed mutagenesis by the CRISPR/Cas9 system removed two major allergenic proteins from mature soybean seeds. This system enables rapid and efficient modification of seed components in soybean varieties.

Mikan Genome Database (MiGD): integrated database of genome annotation, genomic diversity, and CAPS marker information for mandarin molecular breeding.

Citrus species are some of the most valuable and widely consumed fruits globally. The genome sequences of representative citrus (e.g., Citrus clementina, C. sinensis, C. grandis) species have been released but the research base for mandarin molecular breeding is still poor. We assembled the genomes of Citrus unshiu and Poncirus trifoliata, two important species for citrus industry in Japan, using hybrid de novo assembly of Illumina and PacBio sequence data, and developed the Mikan Genome Database (MiGD). The assembled genome sizes of C. unshiu and P. trifoliata are 346 and 292 Mb, respectively, similar to those of citrus species in public databases; they are predicted to possess 41,489 and 34,333 protein-coding genes in their draft genome sequences, with 9,642 and 8,377 specific genes when compared to C. clementina, respectively. MiGD is an integrated database of genome annotation, genetic diversity, and Cleaved Amplified Polymorphic Sequence (CAPS) marker information, with these contents being mutually linked by genes. MiGD facilitates access to genome sequences of interest from previously reported linkage maps through CAPS markers and obtains polymorphism information through the multiple genome browser TASUKE. The genomic resources in MiGD (https://mikan.dna.affrc.go.jp) could provide valuable information for mandarin molecular breeding in Japan.

Spatial and temporal adaptation of predictive saccades based on motion inference.

Moving objects are often occluded behind larger, stationary objects, but we can easily predict when and where they reappear. Here, we show that the prediction of object reappearance is subject to adaptive learning. When monkeys generated predictive saccades to the location of target reappearance, systematic changes in the location or timing of target reappearance independently altered the endpoint or latency of the saccades. Furthermore, spatial adaptation of predictive saccades did not alter visually triggered reactive saccades, whereas adaptation of reactive saccades altered the metrics of predictive saccades. Our results suggest that the extrapolation of motion trajectory may be subject to spatial and temporal recalibration mechanisms located upstream from the site of reactive saccade adaptation. Repetitive exposure of visual error for saccades induces qualitatively different adaptation, which might be attributable to different regions in the cerebellum that regulate learning of trajectory prediction and saccades.

Foreign DNA detection by high-throughput sequencing to regulate genome-edited agricultural products.

Although the advent of several new breeding techniques (NBTs) is revolutionizing agricultural production processes, technical information necessary for their regulation is yet to be provided. Here, we show that high-throughput DNA sequencing is effective for the detection of unintended remaining foreign DNA segments in genome-edited rice. A simple k-mer detection method is presented and validated through a series of computer simulations and real data analyses. The data show that a short foreign DNA segment of 20 nucleotides can be detected and the probability that the segment is overlooked is 10-3 or less if the average sequencing depth is 30 or more, while the number of false hits is less than 1 on average. This method was applied to real sequencing data, and the presence and absence of an external DNA segment were successfully proven. Additionally, our in-depth analyses also identified some weaknesses in current DNA sequencing technologies. Hence, for a rigorous safety assessment, the combination of k-mer detection and another method, such as Southern blot assay, is recommended. The results presented in this study will lay the foundation for the regulation of NBT products, where foreign DNA is utilized during their generation.

TASUKE+: a web-based platform for exploring genome-wide association studies results and large-scale resequencing data.

Recent revolutionary advancements in sequencing technologies have made it possible to obtain mass quantities of genome-scale sequence data in a cost-effective manner and have drastically altered molecular biological studies. To utilize these sequence data, genome-wide association studies (GWASs) have become increasingly important. Hence, there is an urgent need to develop a visualization tool that enables efficient data retrieval, integration of GWAS results with diverse information and rapid public release of such large-scale genotypic and phenotypic data. We developed a web-based genome browser TASUKE+ (https://tasuke.dna.affrc.go.jp/), which is equipped with the following functions: (i) interactive GWAS results visualization with genome resequencing data and annotation information, (ii) PCR primer design, (iii) phylogenetic tree reconstruction and (iv) data sharing via the web. GWAS results can be displayed in parallel with polymorphism data, read depths and annotation information in an interactive and scalable manner. Users can design PCR primers for polymorphic sites of interest. In addition, a molecular phylogenetic tree of any region can be reconstructed so that the overall relationship among the examined genomes can be understood intuitively at a glance. All functions are implemented through user-friendly web-based interfaces so that researchers can easily share data with collaborators in remote places without extensive bioinformatics knowledge.

Genome editing in plants by engineered CRISPR-Cas9 recognizing NG PAM.

Streptococcus pyogenes Cas9 (SpCas9) is widely used for genome editing and requires NGG as a protospacer adjacent motif (PAM). Here, we show that the engineered SpCas9 (SpCas9-NGv1) can efficiently mutagenize endogenous target sites with NG PAMs in the rice and Arabidopsis genomes. Furthermore, we demonstrate that the SpCas9-NGv1 nickase fused to cytidine deaminase mediates C-to-T substitutions near the 5' end of the target sequence.

The Spleen Is an Ideal Site for Inducing Transplanted Islet Graft Expansion in Mice.

Alternative islet transplantation sites have the potential to reduce the marginal number of islets required to ameliorate hyperglycemia in recipients with diabetes. Previously, we reported that T cell leukemia homeobox 1 (Tlx1)+ stem cells in the spleen effectively regenerated into insulin-producing cells in the pancreas of non-obese diabetic mice with end-stage disease. Thus, we investigated the spleen as a potential alternative islet transplantation site. Streptozotocin-induced diabetic C57BL/6 mice received syngeneic islets into the portal vein (PV), beneath the kidney capsule (KC), or into the spleen (SP). The marginal number of islets by PV, KC, or SP was 200, 100, and 50, respectively. Some plasma inflammatory cytokine levels in the SP group were significantly lower than those of the PV group after receiving a marginal number of islets, indicating reduced inflammation in the SP group. Insulin contents were increased 280 days after islet transplantation compared with those immediately following transplantation (p<0.05). Additionally, Tlx1-related genes, including Rrm2b and Pla2g2d, were up-regulated, which indicates that islet grafts expanded in the spleen. The spleen is an ideal candidate for an alternative islet transplantation site because of the resulting reduced inflammation and expansion of the islet graft.

Genomic dissection of a 'Fuji' apple cultivar: re-sequencing, SNP marker development, definition of haplotypes, and QTL detection.

'Fuji' is one of the most popular and highly-produced apple cultivars worldwide, and has been frequently used in breeding programs. The development of genotypic markers for the preferable phenotypes of 'Fuji' is required. Here, we aimed to define the haplotypes of 'Fuji' and find associations between haplotypes and phenotypes of five traits (harvest day, fruit weight, acidity, degree of watercore, and flesh mealiness) by using 115 accessions related to 'Fuji'. Through the re-sequencing of 'Fuji' genome, total of 2,820,759 variants, including single nucleotide polymorphisms (SNPs) and insertions or deletions (indels) were detected between 'Fuji' and 'Golden Delicious' reference genome. We selected mapping-validated 1,014 SNPs, most of which were heterozygous in 'Fuji' and capable of distinguishing alleles inherited from the parents of 'Fuji' (i.e., 'Ralls Janet' and 'Delicious'). We used these SNPs to define the haplotypes of 'Fuji' and trace their inheritance in relatives, which were shown to have an average of 27% of 'Fuji' genome. Analysis of variance (ANOVA) based on 'Fuji' haplotypes identified one quantitative trait loci (QTL) each for harvest time, acidity, degree of watercore, and mealiness. A haplotype from 'Delicious' chr14 was considered to dominantly cause watercore, and one from 'Ralls Janet' chr1 was related to low-mealiness.

Pretreatment of donor islets with papain improves allograft survival without systemic immunosuppression in mice.

Although current immunosuppression protocols improve the efficacy of clinical allogenic islet transplantation, T cell-mediated allorejection remains unresolved, and major histocompatibility complexes (MHCs) play a crucial role in this process. Papain, a cysteine protease, has the unique ability to cleave the extracellular domain of the MHC class I structure. We hypothesized that pretreatment of donor islets with papain would diminish the expression of MHC class I on islets, reducing allograft immunogenicity and contributing to prolongation of islet allograft survival. BALB/c islets pretreated with papain were transplanted into C57BL/6J mice as an acute allorejection model. Treatment with 1 mg/mL papain significantly prolonged islet allograft survival. In vitro, to determine the inhibitory effect on T cell-mediated alloreactions, we performed lymphocyte proliferation assays and mixed lymphocyte reactions. Host T cell activation against allogenic islet cells was remarkably suppressed by pretreatment of donor islet cells with 10 mg/mL papain. Flow cytometric analysis was also performed to investigate the effect of papain treatment on the expression of MHC class I on islets. One or 10 mg/mL papain treatment reduced MHC class I expression on the islet cell surface. Pretreatment of donor islets with papain suppresses MHC class I-mediated allograft rejection in mice and contributes to prolongation of islet allograft survival without administration of systemic immunosuppressants. These results suggest that pretreatment of human donor islets with papain may reduce the immunogenicity of the donor islets and minimize the dosage of systemic immunosuppressants required in a clinical setting.

Progress in Clinical Encapsulated Islet Xenotransplantation.

At the 2015 combined congress of the Cell Transplant Society, International Pancreas and Islet Transplant Association, and International Xenotransplantation Association, a symposium was held to discuss recent progress in pig islet xenotransplantation. The presentations focused on 5 major topics - (1) the results of 2 recent clinical trials of encapsulated pig islet transplantation, (2) the inflammatory response to encapsulated pig islets, (3) methods to improve the secretion of insulin by pig islets, (4) genetic modifications to the islet-source pigs aimed to protect the islets from the primate immune and/or inflammatory responses, and (5) regulatory aspects of clinical pig islet xenotransplantation. Trials of microencapsulated porcine islet transplantation to treat unstable type 1 diabetic patients have been associated with encouraging preliminary results. Further advances to improve efficacy may include (1) transplantation into a site other than the peritoneal cavity, which might result in better access to blood, oxygen, and nutrients; (2) the development of a more biocompatible capsule and/or the minimization of a foreign body reaction; (3) pig genetic modification to induce a greater secretion of insulin by the islets, and/or to reduce the immune response to islets released from damaged capsules; and (4) reduction of the inflammatory response to the capsules/islets by improvements in the structure of the capsules and/or in genetic engineering of the pigs and/or in some form of drug therapy. Ethical and regulatory frameworks for islet xenotransplantation are already available in several countries, and there is now a wider international perception of the importance of developing an internationally harmonized ethical and regulatory framework.

The Nipponbare genome and the next-generation of rice genomics research in Japan.

The map-based genome sequence of the japonica rice cultivar Nipponbare remains to date as the only monocot genome that has been sequenced to a high-quality level. It has become the reference sequence for understanding the diversity among thousands of rice cultivars and its wild relatives as well as the major cereal crops that comprised the food source for the entire human race. This review focuses on the accomplishments in rice genomics in Japan encompassing the last 10 years which have led into deeper understanding of the genome, characterization of many agronomic traits, comprehensive analysis of the transcriptome, and the map-based cloning of many genes associated with agronomic traits.

Islet-derived damage-associated molecular pattern molecule contributes to immune responses following microencapsulated neonatal porcine islet xenotransplantation in mice.

BACKGROUND: Clinical allogeneic islet transplantation has become an attractive procedure for type 1 diabetes mellitus treatment. However, there is a severe shortage of human donors. Microencapsulated neonatal porcine islet (NPI) xenotransplantation may be an alternative transplantation procedure. Currently, the efficacy of microencapsulated NPI xenotransplantation into the peritoneal cavity is limited because of early non-function resulting from inflammation, which is a serious hindrance to promoting this procedure as a standard therapy. Previously, we have demonstrated that high-mobility group box 1 (HMGB1), a damage-associated molecular pattern (DAMP) molecule, was released from transplanted islets and triggered inflammatory reactions leading to early loss of intrahepatic syngeneic islet grafts in mice. In this study, we hypothesized that the inflammatory reaction in the peritoneal cavity following the transplantation of microencapsulated NPIs is more severe than that of empty capsules. Additionally, we predicted that HMGB1 released from transplanted microencapsulated NPIs triggers further inflammatory reactions in mice. Finally, we hypothesized that microencapsulated NPI xenotransplantation efficacy would be improved by treatment-targeting inflammatory reactions in a mouse model. METHODS: A total of 10 000 empty capsules (alginate-poly-L-ornithine-alginate) or 10 000 IEQ microencapsulated NPIs were transplanted into the peritoneal cavity of streptozotocin-induced diabetic C57BL/6 mice. RESULTS: The numbers of mononuclear cells in the peritoneal cavity following empty capsule or microencapsulated NPI transplantation were 4.8 × 10(6)  ± 0.9 × 10(6) and 13.6 × 10(6)  ± 3.0 × 10(6) , respectively (P < 0.05). Fluorescence-activated cell sorting (FACS) analysis revealed that tumor necrosis factor (TNF)-α-, interleukin (IL)-6-, interferon (IFN)-γ-, and/or IL-12-positive macrophages, neutrophils, and dendritic cells had infiltrated the peritoneal cavity after empty capsules or microencapsulated NPIs administration. IL-6 concentrations in the peritoneal lavage fluids on 7 days after empty capsule or microencapsulated NPI transplantation were 18.5 ± 10.0 and 157.4 ± 46.3 pg/ml, respectively (P < 0.001), while TNF-α concentrations were 4.6 ± 1.4 and 19.8 ± 8.4 pg/ml, respectively (P < 0.01). In addition, HMGB1 concentrations were 37.6 ± 6.6 and 117.4 ± 8.1 ng/ml, respectively (P < 0.0001). In vitro experiments revealed that the total amount of released HMGB1 into the culture medium of empty capsule (200 capsules/dish) and microencapsulated NPI (200 IEQ/dish) after hypoxic culture (1% O2 , 5% CO2 , and 94% N2 ) was 0 and 8.6 ± 2.2 ng, respectively (P < 0.001). FACS analysis revealed that TNF-α- and IL-6-positive macrophages were also observed in the peritoneal cavity following intraperitoneal injection of HMGB1 itself. Anti-TNF-α antibody treatment was associated with slightly prolonged graft survival and improved glucose tolerance 30 days after transplantation, but none of the recipients were remained normoglycemic. CONCLUSIONS: In conclusion, early inflammatory reactions might be therapeutic targets for the prolongation of microencapsulated NPIs graft survival. Thus, treatment-targeting inflammation might improve the efficiency of clinical microencapsulated NPI xenotransplantation.

TENOR: Database for Comprehensive mRNA-Seq Experiments in Rice.

Here we present TENOR (Transcriptome ENcyclopedia Of Rice, http://tenor.dna.affrc.go.jp), a database that encompasses large-scale mRNA sequencing (mRNA-Seq) data obtained from rice under a wide variety of conditions. Since the elucidation of the ability of plants to adapt to various growing conditions is a key issue in plant sciences, it is of great interest to understand the regulatory networks of genes responsible for environmental changes. We used mRNA-Seq and performed a time-course transcriptome analysis of rice, Oryza sativa L. (cv. Nipponbare), under 10 abiotic stress conditions (high salinity; high and low phosphate; high, low and extremely low cadmium; drought; osmotic; cold; and flood) and two plant hormone treatment conditions (ABA and jasmonic acid). A large number of genes that were responsive to abiotic stresses and plant hormones were detected by differential expression analysis. Furthermore, several responsive genes were found to encode transcription factors that could control the transcriptional network of stress responses, but the timing of the induction of these genes was not uniform across conditions. A significant number of cis-regulatory elements were enriched in the promoter regions of the responsive genes and were shared among conditions. These data suggest that some key components of gene regulation networks are shared between different stress signaling pathways. All the resources (novel genes identified from mRNA-Seq data, expression profiles, co-expressed genes and cis-regulatory elements) can be searched for and are available in TENOR.

Improvement of barley genome annotations by deciphering the Haruna Nijo genome.

Full-length (FL) cDNA sequences provide the most reliable evidence for the presence of genes in genomes. In this report, detailed gene structures of barley, whole genome shotgun (WGS) and additional transcript data of the cultivar Haruna Nijo were quality controlled and compared with the published Morex genome information. Haruna Nijo scaffolds have longer total sequence length with much higher N50 and fewer sequences than those in Morex WGS contigs. The longer Haruna Nijo scaffolds provided efficient FLcDNA mapping, resulting in high coverage and detection of the transcription start sites. In combination with FLcDNAs and RNA-Seq data from four different tissue samples of Haruna Nijo, we identified 51,249 gene models on 30,606 loci. Overall sequence similarity between Haruna Nijo and Morex genome was 95.99%, while that of exon regions was higher (99.71%). These sequence and annotation data of Haruna Nijo are combined with Morex genome data and released from a genome browser. The genome sequence of Haruna Nijo may provide detailed gene structures in addition to the current Morex barley genome information.