The outbreak of novel coronavirus (COVID-19) infections occurring in 2019 is in dire need of finding potential therapeutic agents. In this study, we used molecular docking strategies to repurpose HIV protease inhibitors and nucleotide analogues for COVID-19. The evaluation was made on docking scores calculated by AutoDock Vina and RosettaCommons. Preliminary results suggested that Indinavir and Remdesivir have the best docking scores and the comparison of the docking sites of these two drugs shows a near perfect dock in the overlap region of the protein pocket. However, the active sites inferred from the proteins of SARS coronavirus are not compatible with the docking site of COVID-19, which may give rise to concern in the efficacy of drugs.

Background Cluster headache is a highly debilitating neurological disorder with considerable inter-ethnic differences. Genome-wide association studies (GWAS) recently identified replicable genomic loci for cluster headache in Europeans, but the genetic underpinnings for cluster headache in Asians remain unclear. The objective of this study is to investigate the genetic architecture and susceptibility loci of cluster headache in Han Chinese resided in Taiwan. Methods We conducted a two-stage genome-wide association study in a Taiwanese cohort enrolled from 2007 through 2022 to identify the genetic variants associated with cluster headache. Diagnosis of cluster headache was retrospectively ascertained with the criteria of International Classification of Headache Disorders, third edition. Control subjects were enrolled from the Taiwan Biobank. Genotyping was conducted with the Axiom Genome-Wide Array TWB chip, followed by whole genome imputation. A polygenic risk score was developed to differentiate patients from controls. Downstream analyses including gene-set and tissue enrichment, linkage disequilibrium score regression, and pathway analyses were performed. Results We enrolled 734 patients with cluster headache and 9,846 population-based controls. We identified three replicable loci, with the lead SNPs being rs1556780 in CAPN2 (odds ratio = 1.59, 95% CI 1.42‒1.78, p = 7.61 × 10–16), rs10188640 in MERTK (odds ratio = 1.52, 95% CI 1.33‒1.73, p = 8.58 × 10–13), and rs13028839 in STAB2 (odds ratio = 0.63, 95% CI 0.52‒0.78, p = 2.81 × 10–8), with the latter two replicating the findings in European populations. Several previously reported genes also showed significant associations with cluster headache in our samples. Polygenic risk score differentiated patients from controls with an area under the receiver operating characteristic curve of 0.77. Downstream analyses implicated circadian regulation and immunological processes in the pathogenesis of cluster headache. Conclusions This study revealed the genetic architecture and novel susceptible loci of cluster headache in Han Chinese residing in Taiwan. Our findings support the common genetic contributions of cluster headache across ethnicities and provide novel mechanistic insights into the pathogenesis of cluster headache.

The selection of peptides presented by MHC molecules is crucial for antigen discovery. Previously, several predictors have shown impressive performance on binding affinity. However, the decisive MHC residues and their relation to the selection of binding peptides are still unrevealed. Here, we connected HLA alleles with binding motifs via our deep learning-based framework, MHCfovea. MHCfovea expanded the knowledge of MHC-I-binding motifs from 150 to 13,008 alleles. After clustering N-terminal and C-terminal sub-motifs on both observed and unobserved alleles, MHCfovea calculated the hyper-motifs and the corresponding allele signatures on the important positions to disclose the relation between binding motifs and MHC-I sequences. MHCfovea delivered 32 pairs of hyper-motifs and allele signatures (HLA-A: 13, HLA-B: 12, and HLA-C: 7). The paired hyper-motifs and allele signatures disclosed the critical polymorphic residues that determine the binding preference, which are believed to be valuable for antigen discovery and vaccine design when allele specificity is concerned.

Motivation: Whole genome sequencing (WGS) by next-generation sequencing produces millions of variants for an individual. The retrieval of biomedical literature for such a large number of genetic variants remains challenging, because in many cases the variants are only present in tables as images, or in the supplementary documents of which the file formats are diverse. Results: The proposed tool named variant2literature from the TaiGenomics (Toolkits for AI genomics) resolves the problem by incorporating text recognition with image processing. In addition to the adoption of advanced text retrieval, the recall rate of finding the literature containing the variants of interest is further improved by employing the skill of variant normalization. Different variant presentations are transformed into chromosome coordinates (standard VCF format) such that false negatives can be largely avoided. variant2literature is available in two ways. First, a web-based interface is provided to search all the literature in PMC Open Access Subset. Second, the command-line executable can be downloaded such that the users are free to search all the files in a specified directory locally.

Understanding the genetic basis of human complex diseases is increasingly important in the development of precision medicine. Over the last decade, genome-wide association studies (GWAS) have become a key technique for detecting associations between common diseases and single nucleotide polymorphisms (SNPs) present in a cohort of individuals. Alternatively, the polygenic risk score (PRS), which often applies results from GWAS summary statistics, is calculated for the estimation of genetic propensity to a trait at the individual level. Despite many GWAS and PRS tools being available to analyze a large volume of genotype data, most clinicians and medical researchers are often not familiar with the bioinformatics tools and lack access to a high-performance computing cluster resource. To fill this gap, we provide a publicly available web server, PGSbuilder, for the GWAS and PRS analysis of human genomes with variant annotations. The user-friendly and intuitive PGSbuilder web server is developed to facilitate the discovery of the genetic variants associated with complex traits and diseases for medical professionals with limited computational skills. For GWAS analysis, PGSbuilder provides the most renowned analysis tool PLINK 2.0 package. For PRS, PGSbuilder provides six different PRS methods including Clumping and Thresholding, Lassosum, LDPred2, GenEpi, PRS-CS, and PRSice2. Furthermore, PGSbuilder provides an intuitive user interface to examine the annotated functional effects of variants from known biomedical databases and relevant literature using advanced natural language processing approaches. In conclusion, PGSbuilder offers a reliable platform to aid researchers in advancing the public perception of genomic risk and precision medicine for human disease genetics. PGSbuilder is freely accessible at http://pgsb.tw23.org.

The selection of peptides presented by MHC molecules is crucial for antigen discovery. Previously, several predictors have shown impressive performance on binding affinity. However, the decisive MHC residues and their relation to the selection of binding peptides are still unrevealed. Here, we connected HLA alleles with binding motifs via our deep learning-based framework, MHCfovea. MHCfovea expanded the knowledge of MHC-I-binding motifs from 150 to 13,008 alleles. After clustering N-terminal and C-terminal sub-motifs on both observed and unobserved alleles, MHCfovea calculated the hyper-motifs and the corresponding allele signatures on the important positions to disclose the relation between binding motifs and MHC-I sequences. MHCfovea delivered 32 pairs of hyper-motifs and allele signatures (HLA-A: 13, HLA-B: 12, and HLA-C: 7). The paired hyper-motifs and allele signatures disclosed the critical polymorphic residues that determine the binding preference, which are believed to be valuable for antigen discovery and vaccine design when allele specificity is concerned.

e13537 Background: Cell-free miRNAs (cf-miRNA), circulated in body fluids such as plasma or serum, have shown their ability to detect, diagnose, and monitor cancers. Combining machine learning (ML) technology with these biomarkers facilitates early detection of cancers, which increases the accuracy of clinical decisions and empowers people to take control of their health status. However, the data of the cf-miRNAs has characteristics, which will affect the results of ML. Therefore, this study tries to expound on them in different aspects and to build a reasonable model. Methods: We downloaded large-scale datasets of the platform GPL21263 from the Gene Expression Omnibus for modeling experiments. We curated 8,174 subjects with 2,565 miRNA targets across 7 cancer types of different cf-miRNA-based cancer studies. Moreover, we used principal component analysis (PCA) to observe the datasets, recursive feature elimination (RFE) for feature selection, and tree-based algorithms to build the prediction model. Results: The characteristics of the cf-miRNA we like to share are: (1) Cancer subjects express more cf-miRNAs than control subjects. In the control group, there were 294 and 327 miRNAs with missing rates under 50% and 25%, respectively. In contrast, there were 395 and 485 miRNAs with the same thresholds in the cancer group; (2) Dividing subjects into cancer and control groups is simpler than distinguishing specific cancer types. In the PCA, the average Euclidean distance between the control group and each cancer type is 98.48, while it is 20.23 within each cancer type; (3) For obtaining cancer-specific biomarkers, we suggested that other non-target cancer subjects should be considered as negative controls. We modeled 7 cancer types and compared the proportion of cancer-specific biomarkers, not selected by any other models. The proportion increased from 30.0% to 57.8% after we added other non-target cancer subjects to the control group; next, we focus on multi-cancer modeling: (4) We need at least 400 samples to distinguish seven cancer types. In our experiment, we kept increasing the size of the training data, a hundred at a time. As data was added to the model, the accuracy increased but plateaued after adding 400 samples; (5) Based on RFE, 120 miRNAs is a reasonable number to distinguish seven cancers. Moreover, we found some of these miRNAs are only expressed in cancer subjects. We might lose this kind of biomarkers if we filtered out them by the missing rate; (6) The 10-fold cross validation accuracy of the multi-cancer model can achieve 93.0% using the gradient-boosted trees algorithm. Conclusions: In this study, we showed the guidance for modeling miRNAs in different aspects, including labeling strategy, sample and feature sizes, and the high-accuracy multi-cancer model we can achieve. We hope this guidance would inspire researchers on cf-miRNA-related machine learning applications.

No abstract