Search

We aimed to investigate the value of computed tomography (CT)-based radiomics with artificial intelligence (AI) in predicting pathological lymph node metastasis (pN) in patients with clinical stage 0–IA non-small cell lung cancer (c-stage 0–IA NSCLC). This study enrolled 720 patients who underwent complete surgical resection for c-stage 0–IA NSCLC, and were assigned to the derivation and validation cohorts. Using the AI software Beta Version (Fujifilm Corporation, Japan), 39 AI imaging factors, including 17 factors from the AI ground-glass nodule analysis and 22 radiomics features from nodule characterization analysis, were extracted to identify factors associated with pN. Multivariate analysis showed that clinical stage IA3 (p = 0.028), solid-part size (p p = 0.033) were independently associated with pN. The receiver operating characteristic analysis showed that the area under the curve and optimal cut-off values of the average solid CT value relevant to pN were 0.761 and -103 Hounsfield units, and the threshold provided sensitivity, specificity, and negative predictive values of 69%, 65%, and 94% in the entire cohort, respectively. Measuring the average solid-CT value of tumors for pN may have broad applications such as guiding individualized surgical approaches and postoperative treatment.

Laboratory mouse strains have mosaic genomes derived from at least three major subspecies that are distributed in Eurasia. Here, we describe genomic variations in ten inbred strains: Mus musculus musculus-derived BLG2/Ms, NJL/Ms, CHD/Ms, SWN/Ms, and KJR/Ms; M. m. domesticus-derived PGN2/Ms and BFM/Ms; M. m. castaneus-derived HMI/Ms; and JF1/Ms and MSM/Ms, which were derived from a hybrid between M. m. musculus and M. m. castaneus. These strains were established by Prof. Moriwaki in the 1980s and are collectively named the “Mishima Battery”. These strains show large phenotypic variations in body size and in many physiological traits. We resequenced the genomes of the Mishima Battery strains and performed a comparative genomic analysis with dbSNP data. More than 81 million nucleotide coordinates were identified as variant sites due to the large genetic distances among the mouse subspecies; 8,062,070 new SNP sites were detected in this study, and these may underlie the large phenotypic diversity observed in the Mishima Battery. The new information was collected in a reconstructed genome database, termed MoG+ that includes new application software and viewers. MoG+ intuitively visualizes nucleotide variants in genes and intergenic regions, and amino acid substitutions across the three mouse subspecies. We report statistical data from the resequencing and comparative genomic analyses and newly collected phenotype data of the Mishima Battery, and provide a brief description of the functions of MoG+, which provides a searchable and unique data resource of the numerous genomic variations across the three mouse subspecies. The data in MoG+ will be invaluable for research into phenotype-genotype links in diverse mouse strains.

Although domesticated goldfish strains exhibit highly diversified phenotypes in morphology, the genetic basis underlying these phenotypes is poorly understood. Here, based on analysis of transposable elements in the allotetraploid goldfish genome, we found that its two subgenomes have evolved asymmetrically since a whole-genome duplication event in the ancestor of goldfish and common carp. We conducted whole-genome sequencing of 27 domesticated goldfish strains and wild goldfish. We identified more than 60 million genetic variations and established a population genetic structure of major goldfish strains. Genome-wide association studies and analysis of strain-specific variants revealed genetic loci associated with several goldfish phenotypes, including dorsal fin loss, long-tail, telescope-eye, albinism, and heart-shaped tail. Our results suggest that accumulated mutations in the asymmetrically evolved subgenomes led to generation of diverse phenotypes in the goldfish domestication history. This study is a key resource for understanding the genetic basis of phenotypic diversity among goldfish strains.

Mutations generate genetic variation and are a major driving force of evolution. Therefore, examining mutation rates and modes are essential for understanding the genetic basis of the physiology and evolution of organisms. Here, we aim to identify germline de novo mutations through the whole-genome surveyance of Mendelian inheritance error sites (MIEs), those not inherited through the Mendelian inheritance manner from either of the parents, using ultra-deep whole genome sequences (>150-fold) from a chimpanzee parent-offspring trio. We identified such 889 MIEs and classified them into four categories based on the pattern of inheritance and the sequence read depth: [i] de novo single nucleotide variants (SNVs), [ii] copy number neutral inherited variants, [iii] hemizygous deletion inherited variants, and [iv] de novo copy number variants (CNVs). From de novo SNV candidates, we estimated a germline de novo SNV mutation rate as 1.48 × 10−8 per site per generation or 0.62 × 10−9 per site per year. In summary, this study demonstrates the significance of ultra-deep whole genome sequencing not only for the direct estimation of mutation rates but also for discerning various mutation modes including de novo allelic conversion and de novo CNVs by identifying MIEs through the transmission of genomes from parents to offspring., チンパンジー親子トリオ（父親－母親－息子）の 全ゲノム配列を高精度で解明. 京都大学プレスリリース. 2017-11-02.

The design of graphical Kansei models varies among Kansei domains. This report introduces a Kansei modeling technique based on a Bayesian network and correlation analysis. These methods can explain the connection among concepts such as human physiological and psychological responses in a simple structure. This technique is expected to be helpful to start Kansei modeling for new domains with basic knowledge of Kansei and to construct a graphical Bayesian model to explain undefined Kansei factors of a product. For this study, we use experimental data and questionnaire results of facial pack application to describe this technique. We intend to show how Bayesian networks can clarify the effectiveness of a specific mask for individual users.

The branching pattern of pulmonary veins exhibits many variations. Here we describe 2 patients with lung malignancies who were found to have pulmonary vein anomalies. We performed video-assisted thoracoscopic lobectomies based on preoperative simulations by 3-dimensional computed tomography. Using 3-dimensional computed tomography, a surgeon can easily construct digital images of the patient’s pulmonary vessels within a few minutes. Simulation using 3-dimensional computed tomography imaging is useful for further understanding of the surgical anatomy.

Preventing, diagnosing, and treating disease is greatly facilitated by the availability of biomarkers. Recent improvements in bioinformatics technology have facilitated large-scale screening of DNA microarrays for candidate biomarkers. Here we discuss a gene selection method, which is called LEAve-one-out Forward selection method (LEAF), for discovering informative genes embedded in gene expression data, and propose an additional algorithm for extending LEAF’s capabilities. LEAF is an iterative forward selection method incorporating the concept of leave-one-out cross validation (LOOCV) and provides a discrimination power score (DPS) for genes, which is a criterion for selecting the candidate of informative genes. We show that LEAF identifies genes that are practically used as biomarkers. Our method should be useful bioinformatics tool for biomedical, clinical, and pharmaceutical researchers.

A biomarker is a substance that is used as a biological parameter of pathological or psychological conditions. Recent progress of bioinformatics technology has enabled large-scale screening of biomarker candidates. As described in this paper, we introduce a gene analysis method, LEAF, for finding informative genes from gene expression data, and propose an additional algorithm for extending LEAF. An iterative forward selection method incorporating the concept of LOOCV, LEAF also provides a discrimination power score (DPS) of genes. Our method shows that the biological functions of extracted genes accorded well with past findings. Our method is a useful tool for many researchers engaged in bioinformatics.