Differential expression of glycosyltransferases identified through comprehensive pan-cancer analysis

Despite accumulating evidence supporting a role for glycosylation in cancer progression and prognosis, the complexity of the human glycome and glycoproteome poses many challenges to understanding glycosylation-related events in cancer. In this study, a multifaceted genomics approach was applied to analyze the impact of differential expression of glycosyltransferases (GTs) in 16 cancers. An enzyme list was compiled and curated from numerous resources to create a consensus set of GTs. Resulting enzymes were analyzed for differential expression in cancer, and findings were integrated with experimental evidence from other analyses, including: similarity of healthy expression patterns across orthologous genes, miRNA expression, automatically-mined literature, curation of known cancer biomarkers, N-glycosylation impact, and survival analysis. The resulting list of GTs comprises 222 human enzymes based on annotations from five databases, 84 of which were differentially expressed in more than five cancers, and 14 of which were observed with the same direction of expression change across all implicated cancers. 25 high-value GT candidates were identified by cross-referencing multimodal analysis results, including PYGM, FUT6 and additional fucosyltransferases, several UDP-glucuronosyltransferases, and others, and are suggested for prioritization in future cancer biomarker studies. Relevant findings are available through OncoMX at https://data.oncomx.org, and the overarching pipeline can be used as a framework for similarly analysis across diverse evidence types in cancer. This work is expected to improve the understanding of glycosylation in cancer by transparently defining the space of glycosyltransferase enzymes and harmonizing variable experimental data to enable improved generation of data-driven cancer biomarker hypotheses.

Genes implicated in Caenorhabditis elegans and human health regulate stress resistance and physical abilities in aged Caenorhabditis elegans

Recently, nine Caenorhabditis elegans genes, grouped into two pathways/clusters, were found to be implicated in healthspan in C. elegans and their homologues in humans, based on literature curation, WormBase data mining and bioinformatics analyses. Here, we further validated these genes experimentally in C. elegans . We downregulated the nine genes via RNA interference (RNAi), and their effects on physical function (locomotion in a swim assay) and on physiological function (survival after heat stress) were analysed in aged nematodes. Swim performance was negatively affected by the downregulation of acox-1.1 , pept-1 , pak-2 , gsk-3 and C25G6.3 in worms with advanced age (twelfth day of adulthood) and heat stress resistance was decreased by RNAi targeting of acox-1.1 , daf-22 , cat-4 , pig-1 , pak-2 , gsk-3 and C25G6.3 in moderately (seventh day of adulthood) or advanced aged nematodes. Only one gene, sad-1 , could not be linked to a health-related function in C. elegans with the bioassays we selected. Thus, most of the healthspan genes could be re-confirmed by health measurements in old worms.

Impact of Automatic Query Generation and Quality Recognition Using Deep Learning to Curate Evidence From Biomedical Literature: Empirical Study

Background The quality of health care is continuously improving and is expected to improve further because of the advancement of machine learning and knowledge-based techniques along with innovation and availability of wearable sensors. With these advancements, health care professionals are now becoming more interested and involved in seeking scientific research evidence from external sources for decision making relevant to medical diagnosis, treatments, and prognosis. Not much work has been done to develop methods for unobtrusive and seamless curation of data from the biomedical literature. Objective This study aimed to design a framework that can enable bringing quality publications intelligently to the users’ desk to assist medical practitioners in answering clinical questions and fulfilling their informational needs. Methods The proposed framework consists of methods for efficient biomedical literature curation, including the automatic construction of a well-built question, the recognition of evidence quality by proposing extended quality recognition model (E-QRM), and the ranking and summarization of the extracted evidence. Results Unlike previous works, the proposed framework systematically integrates the echelons of biomedical literature curation by including methods for searching queries, content quality assessments, and ranking and summarization. Using an ensemble approach, our high-impact classifier E-QRM obtained significantly improved accuracy than the existing quality recognition model (1723/1894, 90.97% vs 1462/1894, 77.21%). Conclusions Our proposed methods and evaluation demonstrate the validity and rigorousness of the results, which can be used in different applications, including evidence-based medicine, precision medicine, and medical education.

WormBase: a modern Model Organism Information Resource

WormBase (https://wormbase.org/) is a mature Model Organism Information Resource supporting researchers using the nematode Caenorhabditis elegans as a model system for studies across a broad range of basic biological processes. Toward this mission, WormBase efforts are arranged in three primary facets: curation, user interface and architecture. In this update, we describe progress in each of these three areas. In particular, we discuss the status of literature curation and recently added data, detail new features of the web interface and options for users wishing to conduct data mining workflows, and discuss our efforts to build a robust and scalable architecture by leveraging commercial cloud offerings. We conclude with a description of WormBase's role as a founding member of the nascent Alliance of Genome Resources.

Impact of Automatic Query Generation and Quality Recognition Using Deep Learning to Curate Evidence From Biomedical Literature: Empirical Study (Preprint)

BACKGROUND The quality of health care is continuously improving and is expected to improve further because of the advancement of machine learning and knowledge-based techniques along with innovation and availability of wearable sensors. With these advancements, health care professionals are now becoming more interested and involved in seeking scientific research evidence from external sources for decision making relevant to medical diagnosis, treatments, and prognosis. Not much work has been done to develop methods for unobtrusive and seamless curation of data from the biomedical literature. OBJECTIVE This study aimed to design a framework that can enable bringing quality publications intelligently to the users’ desk to assist medical practitioners in answering clinical questions and fulfilling their informational needs. METHODS The proposed framework consists of methods for efficient biomedical literature curation, including the automatic construction of a well-built question, the recognition of evidence quality by proposing extended quality recognition model (E-QRM), and the ranking and summarization of the extracted evidence. RESULTS Unlike previous works, the proposed framework systematically integrates the echelons of biomedical literature curation by including methods for searching queries, content quality assessments, and ranking and summarization. Using an ensemble approach, our high-impact classifier E-QRM obtained significantly improved accuracy than the existing quality recognition model (1723/1894, 90.97% vs 1462/1894, 77.21%). CONCLUSIONS Our proposed methods and evaluation demonstrate the validity and rigorousness of the results, which can be used in different applications, including evidence-based medicine, precision medicine, and medical education.

Resolving the hypotheticome: annotating M. tuberculosis gene function through bibliomic reconciliation andstructural modeling

Each decade, billions are invested in Tuberculosis (TB) research to further characterize M. tuberculosis pathogenesis. Despite this investment, nearly half of the 4,031 M. tuberculosis protein-coding genes lack descriptive annotation in community databases, due largely to incomplete reconciliation with the literature and a lack of structure-based methods for functional inference. We coin the term “hypotheticome” as the set of genes in an organism without known function. For M. tuberculosis’ hypotheticome, we compiled the set of genes lacking functional assignment in the most frequently used Mycobacteria annotation database through systematic, exhaustive manual literature curation and 3D-protein structure-based inference, and reconciled these annotations with frequented functional databases, creating a comprehensive M. tuberculosis functional knowledge-base. In doing so, we also introduce standard usage of qualifying adjectives based on quantitative measures of certainty with the hope that this approach is adopted in choosing qualifiers for future functional assignments.Through these methods we functionally annotated 41.3% of the M. tuberculosis hypotheticome, and provide insight into its pathogenesis, antibiotic-resistance, and virulence. Processes implicated in the unique lifestyle of M. tuberculosis of long-term persistence and obligate pathogenesis in genotoxic host microenvironments – lipid metabolism, polyketide biosynthesis, and membrane transport and efflux – were overrepresented in our annotation. Our structural similarity approach unturned proteins that appear critical in host-interaction through apparent host mimicry, particularly involving the phagosome and vesicle-mediated transport, as well as putative structural analogs for highly mutable protein classes, including dozens of PE/PPE family proteins which are major players at the host-pathogen interface, and sixteen potential efflux pumps which are integral to M. tuberculosis drug tolerance. Hypotheses drawn from these proteins’ function may help characterize the onset of latency and identify therapeutic targets. A unified annotation is essential for clear communication about M. tuberculosis. These improvements provide the most comprehensive M. tuberculosis genome annotation to date, and the approach presented can be applied to systematically annotate the genome of other organisms. We provide our novel annotations in General Feature Format with Enzyme Commission and Gene Ontology terms for integration into existing annotation frameworks.

Assessing the Gene-Disease Association of 19 Genes with the RASopathies using the ClinGen Gene Curation Framework

The RASopathies are a complex group of diseases regarding phenotype and genetic etiology. The ClinGen RASopathy Expert Panel assessed published and other publicly available evidence supporting the association of 19 genes with RASopathy conditions. Using the semi-quantitative literature curation method developed by the ClinGen Gene Curation Working Group, evidence for each gene was curated and scored for Noonan syndrome, Costello syndrome, cardiofaciocutaneous (CFC) syndrome, Noonan syndrome with multiple lentigines (NSML), and Noonan-like syndrome with loose anagen hair (NS/LAH).The curated evidence supporting each gene-disease relationship was then discussed and approved by the ClinGen RASopathy Expert Panel. Each association’s strength was classified as Definitive, Strong, Moderate, Limited, Disputed, or No Evidence. Eleven genes were classified as definitively associated with at least one RASopathy condition. Two genes classified as strong for association with at least one RASopathy condition while one gene was moderate and three were limited. The RAS EP also refuted the association of two genes for a RASopathy condition. Overall, our results provide a greater understanding of the different gene-disease relationships within the RASopathies and can help guide and direct clinicians, patients and researchers who are identifying variants in individuals with a suspected RASopathyGRANT NUMBERS:Research reported in this publication was supported by the National Human Genome Research Institute (NHGRI) under award number U41HG006834. MZ received support from German Federal Ministry of Education and Research (BMBF): NSEuroNet (FKZ 01GM1602A), GeNeRARe (FKZ 01GM1519A).

Recent advances in large-scale protein interactome mapping

Protein-protein interactions (PPIs) underlie most, if not all, cellular functions. The comprehensive mapping of these complex networks of stable and transient associations thus remains a key goal, both for systems biology-based initiatives (where it can be combined with other ‘omics’ data to gain a better understanding of functional pathways and networks) and for focused biological studies. Despite the significant challenges of such an undertaking, major strides have been made over the past few years. They include improvements in the computation prediction of PPIs and the literature curation of low-throughput studies of specific protein complexes, but also an increase in the deposition of high-quality data from non-biased high-throughput experimental PPI mapping strategies into publicly available databases.