Efficient HLA imputation from sequential SNPs data by transformer.

Human leukocyte antigen (HLA) genes are associated with a variety of diseases, yet the direct typing of HLA alleles is both time-consuming and costly. Consequently, various imputation methods leveraging sequential single nucleotide polymorphisms (SNPs) data have been proposed, employing either statistical or deep learning models, such as the convolutional neural network (CNN)-based model, DEEP*HLA. However, these methods exhibit limited imputation efficiency for infrequent alleles and necessitate a large size of reference dataset. In this context, we have developed a Transformer-based model to HLA allele imputation, named "HLA Reliable IMpuatioN by Transformer (HLARIMNT)" designed to exploit the sequential nature of SNPs data. We evaluated HLARIMNT's performance using two distinct reference panels; Pan-Asian reference panel (n = 530) and Type 1 Diabetes genetics Consortium (T1DGC) reference panel (n = 5225), alongside a combined panel (n = 1060). HLARIMNT demonstrated superior accuracy to DEEP*HLA across several indices, particularly for infrequent alleles. Furthermore, we explored the impact of varying training data sizes on imputation accuracy, finding that HLARIMNT consistently outperformed across all data size. These findings suggest that Transformer-based models can efficiently impute not only HLA types but potentially other gene types from sequential SNPs data.

Chloride electrode composed of ubiquitous elements for high-energy-density all-solid-state sodium-ion batteries.

Inexpensive and safe energy-storage batteries with high energy densities are in high demand (e.g., for electric vehicles and grid-level renewable energy storage). This study focused on using NaFeCl4, comprising ubiquitous elements, as an electrode material for all-solid-state sodium-ion batteries. Monoclinic NaFeCl4, expected to be the most resource-attractive Fe redox material, is also thermodynamically stable. The Fe2+/3+ redox reaction of the monoclinic NaFeCl4 electrode has a higher potential (3.45 V vs. Na/Na+) than conventional oxide electrodes (e.g., Fe2O3 with 1.5 V vs. Na/Na+) because of the noble properties of chlorine. Additionally, NaFeCl4 exhibits unusually high deformability (99% of the relative density of the pellet) upon uniaxial pressing (382 MPa) at 298 K. NaFeCl4 operates at 333 K in an electrode system containing no electrolyte, thereby realizing next-generation all-solid-state batteries with high safety. A high energy density per positive electrode of 281 Wh kg-1 was achieved using only a simple powder press.

Development and validation of a deep-learning model for scoring of radiographic finger joint destruction in rheumatoid arthritis.

OBJECTIVE: The purpose of this research was to develop a deep-learning model to assess radiographic finger joint destruction in RA. METHODS: The model comprises two steps: a joint-detection step and a joint-evaluation step. Among 216 radiographs of 108 patients with RA, 186 radiographs were assigned to the training/validation dataset and 30 to the test dataset. In the training/validation dataset, images of PIP joints, the IP joint of the thumb or MCP joints were manually clipped and scored for joint space narrowing (JSN) and bone erosion by clinicians, and then these images were augmented. As a result, 11 160 images were used to train and validate a deep convolutional neural network for joint evaluation. Three thousand seven hundred and twenty selected images were used to train machine learning for joint detection. These steps were combined as the assessment model for radiographic finger joint destruction. Performance of the model was examined using the test dataset, which was not included in the training/validation process, by comparing the scores assigned by the model and clinicians. RESULTS: The model detected PIP joints, the IP joint of the thumb and MCP joints with a sensitivity of 95.3% and assigned scores for JSN and erosion. Accuracy (percentage of exact agreement) reached 49.3-65.4% for JSN and 70.6-74.1% for erosion. The correlation coefficient between scores by the model and clinicians per image was 0.72-0.88 for JSN and 0.54-0.75 for erosion. CONCLUSION: Image processing with the trained convolutional neural network model is promising to assess radiographs in RA.

iMusta4SLC: Database for the structural property and variations of solute carrier transporters.

Membrane transporter proteins play important roles in transport of nutrients into the cell, in transport of waste out of the cell, in maintenance of homeostasis, and in signal transduction. Solute carrier (SLC) transporter is the superfamily, which has the largest number of genes (>400 in humans) in membrane transporter and consists of 52 families. SLC transporters carry a wide variety of substrates such as amino acids, peptides, saccharides, ions, neurotransmitters, lipids, hormones and related materials. Despite the apparent importance for the substrate transport, the information of sequence variation and three-dimensional structures have not been integrated to the level of providing new knowledge on the relationship to, for instance, diseases. We, therefore, built a new database named iMusta4SLC, which is available at http://cib.cf.ocha.ac.jp/slc/, that connected the data of structural properties and of pathogenic mutations on human SLC transporters. iMusta4SLC helps to investigate the structural features of pathogenic mutations on SLC transporters. With this database, we found that the mutations at the conserved arginine were frequently involved in diseases, and were located at a border between the membrane and the cytoplasm. Especially in SLC families 2 and 22, the conserved residues formed a large cluster at the border. In SLC2A1, one third of the reported pathogenic missense mutations were found in this conserved cluster.

Influenza virus-host interactome screen as a platform for antiviral drug development.

Host factors required for viral replication are ideal drug targets because they are less likely than viral proteins to mutate under drug-mediated selective pressure. Although genome-wide screens have identified host proteins involved in influenza virus replication, limited mechanistic understanding of how these factors affect influenza has hindered potential drug development. We conducted a systematic analysis to identify and validate host factors that associate with influenza virus proteins and affect viral replication. After identifying over 1,000 host factors that coimmunoprecipitate with specific viral proteins, we generated a network of virus-host protein interactions based on the stage of the viral life cycle affected upon host factor downregulation. Using compounds that inhibit these host factors, we validated several proteins, notably Golgi-specific brefeldin A-resistant guanine nucleotide exchange factor 1 (GBF1) and JAK1, as potential antiviral drug targets. Thus, virus-host interactome screens are powerful strategies to identify targetable host factors and guide antiviral drug development.