Identification of Potential Blood-Based Biomarkers for Frailty by Using an Integrative Approach.

INTRODUCTION: Although frailty is a geriatric syndrome that is associated with disability, hospitalization, and mortality, it can be reversible and preventable with the appropriate interventions. Additionally, as the current diagnostic criteria for frailty include only physical, psychological, cognitive, and social measurements, there is a need for promising blood-based molecular biomarkers to aid in the diagnosis of frailty. METHODS: To identify candidate blood-based biomarkers that can enhance current diagnosis of frailty, we conducted a comprehensive analysis of clinical data, messenger RNA-sequencing (RNA-seq), and aging-related factors using a total of 104 older adults aged 65-90 years (61 frail subjects and 43 robust subjects) in a cross-sectional case-control study. RESULTS: We identified two candidate biomarkers of frailty from the clinical data analysis, nine from the RNA-seq analysis, and six from the aging-related factors analysis. By using combinations of the candidate biomarkers and clinical information, we constructed risk prediction models. The best models used combinations that included skeletal muscle mass index measured by dual-energy X-ray absorptiometry (adjusted p = 0.026), GDF15 (adjusted p = 1.46E-03), adiponectin (adjusted p = 0.012), CXCL9 (adjusted p = 0.011), or apelin (adjusted p = 0.020) as the biomarker. These models achieved a high area under the curve of 0.95 in an independent validation cohort (95% confidence interval: 0.79-0.97). Our risk prediction models showed significantly higher areas under the curve than did models constructed using only basic clinical information (Welch's t test p < 0.001). CONCLUSION: All five biomarkers showed statistically significant correlations with components of the frailty diagnostic criteria. We discovered several potential biomarkers for the diagnosis of frailty. Further refinement may lead to their future clinical use.

The HLA-DRB1*09:01-DQB1*03:03 haplotype is associated with the risk for late-onset Alzheimer's disease in APOE [Formula: see text]4-negative Japanese adults.

Late-onset Alzheimer's disease (LOAD) is the most common cause of dementia among those older than 65 years. The onset of LOAD is influenced by neuroinflammation. The human leukocyte antigen (HLA) system is involved in regulating inflammatory responses. Numerous HLA alleles and their haplotypes have shown varying associations with LOAD in diverse populations, yet their impact on the Japanese population remains to be elucidated. Here, we conducted a comprehensive investigation into the associations between LOAD and HLA alleles within the Japanese population. Using whole-genome sequencing (WGS) data from 303 LOAD patients and 1717 cognitively normal (CN) controls, we identified four-digit HLA class I alleles (A, B, and C) and class II alleles (DRB1, DQB1, and DPB1). We found a significant association between the HLA-DRB1*09:01-DQB1*03:03 haplotype and LOAD risk in APOE [Formula: see text]4-negative samples (odds ratio = 1.81, 95% confidence interval = 1.38-2.38, P = 2.03[Formula: see text]). These alleles not only showed distinctive frequencies specific to East Asians but demonstrated a high degree of linkage disequilibrium in APOE [Formula: see text]4-negative samples (r2 = 0.88). Because HLA class II molecules interact with T-cell receptors (TCRs), we explored potential disparities in the diversities of TCR α chain (TRA) and ß chain (TRB) repertoires between APOE [Formula: see text]4-negative LOAD and CN samples. Lower diversity of TRA repertoires was associated with LOAD in APOE [Formula: see text]4-negative samples, irrespective of the HLA DRB1*09:01-DQB1*03:03 haplotype. Our study enhances the understanding of the etiology of LOAD in the Japanese population and provides new insights into the underlying mechanisms of its pathogenesis.

Association between Circulating Levels of CXCL9 and CXCL10 and Physical Frailty in Older Adults.

INTRODUCTION: Dysregulation of pro-inflammatory chemokines is considered a potential mechanism for the development of age-related medical conditions such as frailty. However, evidence linking circulating chemokines with frailty remains lacking. MATERIALS AND METHODS: We performed a case-control study including 48 cases and 48 controls aged 65-90 years, using the National Center for Geriatrics and Gerontology outpatient registry data. Cases were outpatients with physical frailty and low habitual daily activity. Controls were robust outpatients who performed habitual daily activities. The Japanese version of the Cardiovascular Health Study criteria was used to diagnose physical frailty, and the modified Baecke questionnaire was used to evaluate habitual daily activities. Serum CXCL9 and CXCL10 levels were measured using enzyme-linked immunosorbent assay. RESULTS: The median age (interquartile range) in cases and controls was 78 (73-83) and 76 (72-80) years, with the proportions of men were 47.9% and 43.8%, respectively. In the logistic regression model with adjustment for age, sex, and other confounding factors, the multivariable odds ratios (95% confidence intervals) for the highest versus lowest tertile of CXCL9 and CXCL10 levels were 7.90 (1.61-49.80) and 1.61 (0.42-6.30), respectively. However, we did not observe a linear association between CXCL9 levels and physical frailty components. DISCUSSION/CONCLUSION: Our preliminary data exhibit that circulating CXCL9 levels were positively associated with the odds of physical frailty. However, these findings lack evidence of a dose-response relationship between CXCL9 levels and physical frailty components. Further research with a larger sample size is required to confirm these findings.

RNA-Sequencing Analysis Identification of Potential Biomarkers for Diagnosis of Sarcopenia.

Sarcopenia is a geriatric disease associated with increased mortality and disability. Early diagnosis and intervention are required to prevent it. This study investigated biomarkers for sarcopenia by using a combination of comprehensive clinical data and messenger RNA-sequencing (RNA-seq) analysis obtained from peripheral blood mononuclear cells. We enrolled a total of 114 older adults aged 66-94 years (52 sarcopenia diagnosed according to the Asian Working Group for Sarcopenia 2019 consensus and 62 normal older people). We used clinical data which were not included diagnosis criteria of sarcopenia, and stride length showed significance by logistic regression analysis (Bonferroni corrected p = .012, odds ratio = 0.14, 95% confidence interval [CI]: 0.05-0.40). RNA-seq analysis detected 6 differential expressed genes (FAR1, GNL2, HERC5, MRPL47, NUBP2, and S100A11). We also performed gene-set enrichment analysis and detected 2 functional modules (ie, hub genes, MYH9, and FLNA). By using any combination of the 9 candidates and basic information (age and sex), risk-prediction models were constructed. The best model by using a combination of stride length, HERC5, S100A11, and FLNA, achieved a high area under the curve (AUC) of 0.91 in a validation cohort (95% CI: 0.78-0.95). The quantitative PCR results of the 3 genes were consistent with the trend observed in the RNA-seq results. When BMI was added, the model achieved a high AUC of 0.95 (95% CI: 0.84-0.99). We have discovered potential biomarkers for the diagnosis of sarcopenia. Further refinement may lead to their future practical use in clinical use.

Classification and deep-learning-based prediction of Alzheimer disease subtypes by using genomic data.

Late-onset Alzheimer's disease (LOAD) is the most common multifactorial neurodegenerative disease among elderly people. LOAD is heterogeneous, and the symptoms vary among patients. Genome-wide association studies (GWAS) have identified genetic risk factors for LOAD but not for LOAD subtypes. Here, we examined the genetic architecture of LOAD based on Japanese GWAS data from 1947 patients and 2192 cognitively normal controls in a discovery cohort and 847 patients and 2298 controls in an independent validation cohort. Two distinct groups of LOAD patients were identified. One was characterized by major risk genes for developing LOAD (APOC1 and APOC1P1) and immune-related genes (RELB and CBLC). The other was characterized by genes associated with kidney disorders (AXDND1, FBP1, and MIR2278). Subsequent analysis of albumin and hemoglobin values from routine blood test results suggested that impaired kidney function could lead to LOAD pathogenesis. We developed a prediction model for LOAD subtypes using a deep neural network, which achieved an accuracy of 0.694 (2870/4137) in the discovery cohort and 0.687 (2162/3145) in the validation cohort. These findings provide new insights into the pathogenic mechanisms of LOAD.

Identification of potential blood biomarkers for early diagnosis of Alzheimer's disease through immune landscape analysis.

Mild cognitive impairment (MCI) is a clinical precursor of Alzheimer's disease (AD). Recent genetic studies have reported on associations between AD risk genes and immunity. Here, we obtained samples and data from 317 AD, 432 MCI, and 107 cognitively normal (CN) subjects and investigated immune-cell type composition and immune clonal diversity of T-cell receptor (TRA, TRB, TRG, and TRD) and B-cell receptor (IGH, IGK, and IGL) repertoires through bulk RNA sequencing. We found the proportions of plasma cells, γδ T cells, neutrophils, and B cells were significantly different and the diversities of IGH, IGK, and TRA were significantly small with AD progression. We then identified a differentially expressed gene, WDR37, in terms of risk of MCI-to-AD conversion. Our prognosis prediction model using the potential blood-based biomarkers for early AD diagnosis, which combined two immune repertoires (IGK and TRA), WDR37, and clinical information, successfully classified MCI patients into two groups, low and high, in terms of risk of MCI-to-AD conversion (log-rank test P = 2.57e-3). It achieved a concordance index of 0.694 in a discovery cohort and of 0.643 in an independent validation cohort. We believe that further investigation, using larger sample sizes, will lead to practical clinical use in the near future.

Network-based meta-analysis and the candidate gene association studies reveal novel ethnicity-specific variants in MFSD3 and MRPL43 associated with dementia with Lewy bodies.

Dementia with Lewy bodies (DLB) is the second most common form of neurodegenerative dementia in elderly people, following Alzheimer's disease. Only three genes, SNCA (α-synuclein), APOE (apolipoprotein E), and GBA (glucosylceramidase), have been convincingly demonstrated to be associated with DLB. Here, we applied whole-genome sequencing to blood samples from 61 DLB patients and 45 cognitively normal controls. We used accumulation of candidate mutations to detect novel DLB-associated genes. Subsequent single nucleotide polymorphism (SNP) genotyping and association studies in a large number of samples from Japanese individuals revealed novel heterozygous variants in MFSD3 (rs143475431, c.888T>A:p.C296*; n = 5,421, p = 0.00063) and MRPL43 (chr10:102746730, c.241A>C:p.N81H; n = 4,782, p = 0.0029). We further found that the MFSD3 variant increased plasma levels of butyrylcholinesterase (n = 1,206, p = 0.029). We believe that our findings will contribute to the understanding of DLB and provide insight into its pathogenic mechanism for future studies.

Whole-genome sequencing reveals novel ethnicity-specific rare variants associated with Alzheimer's disease.

Alzheimer's disease (AD) is the most common multifactorial neurodegenerative disease among elderly people. Genome-wide association studies (GWAS) have been highly successful in identifying genetic risk factors. However, GWAS investigate common variants, which tend to have small effect sizes, and rare variants with potentially larger phenotypic effects have not been sufficiently investigated. Whole-genome sequencing (WGS) enables us to detect those rare variants. Here, we performed rare-variant association studies by using WGS data from 140 individuals with probable AD and 798 cognitively normal elder controls (CN), as well as single-nucleotide polymorphism genotyping data from an independent large Japanese AD cohort of 1604 AD and 1235 CN subjects. We identified two rare variants as candidates for AD association: a missense variant in OR51G1 (rs146006146, c.815 G > A, p.R272H) and a stop-gain variant in MLKL (rs763812068, c.142 C > T, p.Q48X). Subsequent in vitro functional analysis revealed that the MLKL stop-gain variant can contribute to increases not only in abnormal cells that should die by programmed cell death but do not, but also in the ratio of Aß42 to Aß40. We further detected AD candidate genes through gene-based association tests of rare variants; a network-based meta-analysis using these candidates identified four functionally important hub genes (NCOR2, PLEC, DMD, and NEDD4). Our findings will contribute to the understanding of AD and provide novel insights into its pathogenic mechanisms that can be used in future studies.

A functional variant of SHARPIN confers increased risk of late-onset Alzheimer's disease.

Late-onset Alzheimer's disease (LOAD) is the most common form of dementia, and its pathogenesis is multifactorial. We previously reported a rare functional variant of SHARPIN (rs572750141, NP_112236.3:p.Gly186Arg) that was significantly associated with LOAD. In addition, several recent studies have suggested the potential role of SHARPIN in AD pathogenesis. In this study, we sought to identify additional functional variants of SHARPIN in Japanese population. Six highly deleterious variants of SHARPIN, comprising four missense variants, one frameshift variant, and one stop-gain variant were detected from whole-genome sequencing data for 180 patients with LOAD and 184 with mild cognitive impairment. One of these candidate variants (rs77359862, NP_112236.3:p.Arg274Trp) was significantly associated with an increased risk of LOAD in 5043 LOAD cases and 11984 controls (P = 0.0016, odds ratio = 1.43). Furthermore, this variant SHARPIN showed aberrant cellular localization and reduced the activation of NF-κB, a central mediator of inflammatory and immune responses. Further investigation of the physiologic role of SHARPIN may reveal the mechanism of onset of LOAD.

JAMIR-eQTL: Japanese genome-wide identification of microRNA expression quantitative trait loci across dementia types.

MicroRNAs (miRNAs) are small non-coding RNAs shown to regulate gene expression by binding to complementary transcripts. Genetic variants, including single-nucleotide polymorphisms and short insertions/deletions, contribute to traits and diseases by influencing miRNA expression. However, the association between genetic variation and miRNA expression remains to be elucidated. Here, by using genotype data and miRNA expression data from 3448 Japanese serum samples, we developed a computational pipeline to systematically identify genome-wide miRNA expression quantitative trait loci (miR-eQTLs). Not only did we identify a total of 2487 cis-miR-eQTLs and 3 155 773 trans-miR-eQTLs at a false discovery rate of <0.05 in six dementia types (Alzheimer's disease, dementia with Lewy bodies, vascular dementia, frontotemporal lobar degeneration, normal-pressure hydrocephalus and mild cognitive impairment) and all samples, including those from patients with other types of dementia, but also we examined the commonality and specificity of miR-eQTLs among dementia types. To enable data searching and downloading of these cis- and trans-eQTLs, we developed a user-friendly database named JAMIR-eQTL, publicly available at https://www.jamir-eqtl.org/. This is the first miR-eQTL database designed for dementia types. Our integrative and comprehensive resource will contribute to understanding the genetic basis of miRNA expression as well as to the discovery of deleterious mutations, particularly in dementia studies. Database URL: https://www.jamir-eqtl.org/.

Dementia subtype prediction models constructed by penalized regression methods for multiclass classification using serum microRNA expression data.

There are many subtypes of dementia, and identification of diagnostic biomarkers that are minimally-invasive, low-cost, and efficient is desired. Circulating microRNAs (miRNAs) have recently gained attention as easily accessible and non-invasive biomarkers. We conducted a comprehensive miRNA expression analysis of serum samples from 1348 Japanese dementia patients, composed of four subtypes-Alzheimer's disease (AD), vascular dementia, dementia with Lewy bodies (DLB), and normal pressure hydrocephalus-and 246 control subjects. We used this data to construct dementia subtype prediction models based on penalized regression models with the multiclass classification. We constructed a final prediction model using 46 miRNAs, which classified dementia patients from an independent validation set into four subtypes of dementia. Network analysis of miRNA target genes revealed important hub genes, SRC and CHD3, associated with the AD pathogenesis. Moreover, MCU and CASP3, which are known to be associated with DLB pathogenesis, were identified from our DLB-specific target genes. Our study demonstrates the potential of blood-based biomarkers for use in dementia-subtype prediction models. We believe that further investigation using larger sample sizes will contribute to the accurate classification of subtypes of dementia.

Whole-genome sequencing with long reads reveals complex structure and origin of structural variation in human genetic variations and somatic mutations in cancer.

BACKGROUND: Identification of germline variation and somatic mutations is a major issue in human genetics. However, due to the limitations of DNA sequencing technologies and computational algorithms, our understanding of genetic variation and somatic mutations is far from complete. METHODS: In the present study, we performed whole-genome sequencing using long-read sequencing technology (Oxford Nanopore) for 11 Japanese liver cancers and matched normal samples which were previously sequenced for the International Cancer Genome Consortium (ICGC). We constructed an analysis pipeline for the long-read data and identified germline and somatic structural variations (SVs). RESULTS: In polymorphic germline SVs, our analysis identified 8004 insertions, 6389 deletions, 27 inversions, and 32 intra-chromosomal translocations. By comparing to the chimpanzee genome, we correctly inferred events that caused insertions and deletions and found that most insertions were caused by transposons and Alu is the most predominant source, while other types of insertions, such as tandem duplications and processed pseudogenes, are rare. We inferred mechanisms of deletion generations and found that most non-allelic homolog recombination (NAHR) events were caused by recombination errors in SINEs. Analysis of somatic mutations in liver cancers showed that long reads could detect larger numbers of SVs than a previous short-read study and that mechanisms of cancer SV generation were different from that of germline deletions. CONCLUSIONS: Our analysis provides a comprehensive catalog of polymorphic and somatic SVs, as well as their possible causes. Our software are available at https://github.com/afujimoto/CAMPHOR and https://github.com/afujimoto/CAMPHORsomatic .

Ethnic and trans-ethnic genome-wide association studies identify new loci influencing Japanese Alzheimer's disease risk.

Alzheimer's disease (AD) has no cure, but early detection and risk prediction could allow earlier intervention. Genetic risk factors may differ between ethnic populations. To discover novel susceptibility loci of AD in the Japanese population, we conducted a genome-wide association study (GWAS) with 3962 AD cases and 4074 controls. Out of 4,852,957 genetic markers that passed stringent quality control filters, 134 in nine loci, including APOE and SORL1, were convincingly associated with AD. Lead SNPs located in seven novel loci were genotyped in an independent Japanese AD case-control cohort. The novel locus FAM47E reached genome-wide significance in a meta-analysis of association results. This is the first report associating the FAM47E locus with AD in the Japanese population. A trans-ethnic meta-analysis combining the results of the Japanese data sets with summary statistics from stage 1 data of the International Genomics of Alzheimer's Project identified an additional novel susceptibility locus in OR2B2. Our data highlight the importance of performing GWAS in non-European populations.

Prognosis prediction model for conversion from mild cognitive impairment to Alzheimer's disease created by integrative analysis of multi-omics data.

BACKGROUND: Mild cognitive impairment (MCI) is a precursor to Alzheimer's disease (AD), but not all MCI patients develop AD. Biomarkers for early detection of individuals at high risk for MCI-to-AD conversion are urgently required. METHODS: We used blood-based microRNA expression profiles and genomic data of 197 Japanese MCI patients to construct a prognosis prediction model based on a Cox proportional hazard model. We examined the biological significance of our findings with single nucleotide polymorphism-microRNA pairs (miR-eQTLs) by focusing on the target genes of the miRNAs. We investigated functional modules from the target genes with the occurrence of hub genes though a large-scale protein-protein interaction network analysis. We further examined the expression of the genes in 610 blood samples (271 ADs, 248 MCIs, and 91 cognitively normal elderly subjects [CNs]). RESULTS: The final prediction model, composed of 24 miR-eQTLs and three clinical factors (age, sex, and APOE4 alleles), successfully classified MCI patients into low and high risk of MCI-to-AD conversion (log-rank test P = 3.44 × 10-4 and achieved a concordance index of 0.702 on an independent test set. Four important hub genes associated with AD pathogenesis (SHC1, FOXO1, GSK3B, and PTEN) were identified in a network-based meta-analysis of miR-eQTL target genes. RNA-seq data from 610 blood samples showed statistically significant differences in PTEN expression between MCI and AD and in SHC1 expression between CN and AD (PTEN, P = 0.023; SHC1, P = 0.049). CONCLUSIONS: Our proposed model was demonstrated to be effective in MCI-to-AD conversion prediction. A network-based meta-analysis of miR-eQTL target genes identified important hub genes associated with AD pathogenesis. Accurate prediction of MCI-to-AD conversion would enable earlier intervention for MCI patients at high risk, potentially reducing conversion to AD.

Lower DNA methylation levels in CpG island shores of CR1, CLU, and PICALM in the blood of Japanese Alzheimer's disease patients.

The aim of the present study was to (1) investigate the relationship between late-onset Alzheimer's disease (AD) and DNA methylation levels in six of the top seven AD-associated genes identified through a meta-analysis of recent genome wide association studies, APOE, BIN1, PICALM, CR1, CLU, and ABCA7, in blood, and (2) examine its applicability to the diagnosis of AD. We examined methylation differences at CpG island shores in the six genes using Sanger sequencing, and one of two groups of 48 AD patients and 48 elderly controls was used for a test or replication analysis. We found that methylation levels in three out of the six genes, CR1, CLU, and PICALM, were significantly lower in AD subjects. The combination of CLU methylation levels and the APOE genotype classified AD patients with AUC = 0.84 and 0.80 in the test and replication analyses, respectively. Our study implicates methylation differences at the CpG island shores of AD-associated genes in the onset of AD and suggests their diagnostic value.

Identification of potential blood biomarkers for early diagnosis of Alzheimer's disease through RNA sequencing analysis.

BACKGROUND: With demographic shifts toward older populations, the number of people with dementia is steadily increasing. Alzheimer's disease (AD) is the most common cause of dementia, and no curative treatment is available. The current best strategy is to delay disease progression and to practice early intervention to reduce the number of patients that ultimately develop AD. Therefore, promising novel biomarkers for early diagnosis are urgently required. METHODS: To identify blood-based biomarkers for early diagnosis of AD, we performed RNA sequencing (RNA-seq) analysis of 610 blood samples, representing 271 patients with AD, 91 cognitively normal (CN) adults, and 248 subjects with mild cognitive impairment (MCI). We first estimated cell-type proportions among AD, MCI, and CN samples from the bulk RNA-seq data using CIBERSORT and then examined the differentially expressed genes (DEGs) between AD and CN samples. To gain further insight into the biological functions of the DEGs, we performed gene set enrichment analysis (GSEA) and network-based meta-analysis. RESULTS: In the cell-type distribution analysis, we found a significant association between the proportion of neutrophils and AD prognosis at a false discovery rate (FDR) < 0.05. Furthermore, a similar trend emerged in the results of routine blood tests from a large number of samples (n = 3,099: AD, 1,605; MCI, 994; CN, 500). In addition, GSEA and network-based meta-analysis based on DEGs between AD and CN samples revealed functional modules and important hub genes associated with the pathogenesis of AD. The risk prediction model constructed by using the proportion of neutrophils and the most important hub genes (EEF2 and RPL7) achieved a high AUC of 0.878 in a validation cohort; when further applied to a prospective cohort, the model achieved a high accuracy of 0.727. CONCLUSIONS: Our model was demonstrated to be effective in prospective AD risk prediction. These findings indicate the discovery of potential biomarkers for early diagnosis of AD, and their further improvement may lead to future practical clinical use.

Bigram-PGK: phosphoglycerylation prediction using the technique of bigram probabilities of position specific scoring matrix.

BACKGROUND: The biological process known as post-translational modification (PTM) is a condition whereby proteomes are modified that affects normal cell biology, and hence the pathogenesis. A number of PTMs have been discovered in the recent years and lysine phosphoglycerylation is one of the fairly recent developments. Even with a large number of proteins being sequenced in the post-genomic era, the identification of phosphoglycerylation remains a big challenge due to factors such as cost, time consumption and inefficiency involved in the experimental efforts. To overcome this issue, computational techniques have emerged to accurately identify phosphoglycerylated lysine residues. However, the computational techniques proposed so far hold limitations to correctly predict this covalent modification. RESULTS: We propose a new predictor in this paper called Bigram-PGK which uses evolutionary information of amino acids to try and predict phosphoglycerylated sites. The benchmark dataset which contains experimentally labelled sites is employed for this purpose and profile bigram occurrences is calculated from position specific scoring matrices of amino acids in the protein sequences. The statistical measures of this work, such as sensitivity, specificity, precision, accuracy, Mathews correlation coefficient and area under ROC curve have been reported to be 0.9642, 0.8973, 0.8253, 0.9193, 0.8330, 0.9306, respectively. CONCLUSIONS: The proposed predictor, based on the feature of evolutionary information and support vector machine classifier, has shown great potential to effectively predict phosphoglycerylated and non-phosphoglycerylated lysine residues when compared against the existing predictors. The data and software of this work can be acquired from https://github.com/abelavit/Bigram-PGK.

A comparison of machine learning classifiers for dementia with Lewy bodies using miRNA expression data.

BACKGROUND: Dementia with Lewy bodies (DLB) is the second most common subtype of neurodegenerative dementia in humans following Alzheimer's disease (AD). Present clinical diagnosis of DLB has high specificity and low sensitivity and finding potential biomarkers of prodromal DLB is still challenging. MicroRNAs (miRNAs) have recently received a lot of attention as a source of novel biomarkers. METHODS: In this study, using serum miRNA expression of 478 Japanese individuals, we investigated potential miRNA biomarkers and constructed an optimal risk prediction model based on several machine learning methods: penalized regression, random forest, support vector machine, and gradient boosting decision tree. RESULTS: The final risk prediction model, constructed via a gradient boosting decision tree using 180 miRNAs and two clinical features, achieved an accuracy of 0.829 on an independent test set. We further predicted candidate target genes from the miRNAs. Gene set enrichment analysis of the miRNA target genes revealed 6 functional genes included in the DHA signaling pathway associated with DLB pathology. Two of them were further supported by gene-based association studies using a large number of single nucleotide polymorphism markers (BCL2L1: P = 0.012, PIK3R2: P = 0.021). CONCLUSIONS: Our proposed prediction model provides an effective tool for DLB classification. Also, a gene-based association test of rare variants revealed that BCL2L1 and PIK3R2 were statistically significantly associated with DLB.

DeepInsight: A methodology to transform a non-image data to an image for convolution neural network architecture.

It is critical, but difficult, to catch the small variation in genomic or other kinds of data that differentiates phenotypes or categories. A plethora of data is available, but the information from its genes or elements is spread over arbitrarily, making it challenging to extract relevant details for identification. However, an arrangement of similar genes into clusters makes these differences more accessible and allows for robust identification of hidden mechanisms (e.g. pathways) than dealing with elements individually. Here we propose, DeepInsight, which converts non-image samples into a well-organized image-form. Thereby, the power of convolution neural network (CNN), including GPU utilization, can be realized for non-image samples. Furthermore, DeepInsight enables feature extraction through the application of CNN for non-image samples to seize imperative information and shown promising results. To our knowledge, this is the first work to apply CNN simultaneously on different kinds of non-image datasets: RNA-seq, vowels, text, and artificial.