A knowledge graph to interpret clinical proteomics data

A systematic mapping study of semantic technologies in multi-omics data integration

OBJECTIVE

The integration of multi-omics data is essential for understanding complex biological systems, providing insights beyond single-omics approaches. However, challenges related to data heterogeneity, standardization, and computational scalability persist. This study explores the interdisciplinary application of semantic technologies to enhance data integration, standardization, and analysis in multi-omics research.

METHODS

We performed a systematic mapping study assessing literature from 2014 to 2024, focusing on the utilization of ontologies, knowledge graphs, and graph-based methods for multi-omics integration.

RESULTS

Our findings indicate a growing number of publications in this field, predominantly appearing in high-impact journals. The deployment of semantic technologies has notably improved data visualization, querying, and management, thus enhancing gene and pathway discovery, and providing deeper disease insights and more accurate predictive modeling.

CONCLUSION

The study underscores the significance of semantic technologies in overcoming multi-omics integration challenges. Future research should focus on integrating diverse data types, developing advanced computational tools, and incorporating AI and machine learning to foster personalized medicine applications.

An update on knowledge graphs and their current and potential applications in drug discovery

INTRODUCTION

Knowledge graphs are becoming prominent tools in computational drug discovery. They effectively integrate heterogeneous biomedical data and generate new hypotheses and knowledge.

AREAS COVERED

This article is based on a literature review using Google Scholar and PubMed to retrieve articles on existing knowledge graphs relevant to the drug discovery field. The authors compare the types of entities, relationships, and data sources they encompass. Additionally, the authors provide examples of their use in the drug discovery field and discuss potential strategies for advancing this research area.

EXPERT OPINION

Knowledge graphs are crucial in drug discovery, but their construction leads to challenges in data integration and consistency. Future research should prioritize the standardization of data sources and data modeling. More efforts are needed for the integration in knowledge graphs of diverse data types, such as chemical structures and epigenetic data, to enhance their effectiveness. Additionally, advancements in large language models should be pursued to aid the development of knowledge graphs, provide intuitive querying capabilities for non-expert users, and explain knowledge graphs -derived predictions, thereby making these tools more accessible and their insights more interpretable for a wider audience.

Shale wastewater treatment policies recommended by integrated knowledge graph and probability-based algorithms

Shale wastewater (SWW) has received much attention recently due to its non-ignorable threat to the environment and human health. Selection of an appropriate SWW treatment technology or a set of technology combinations could be a challenge particularly when no substantial professional-knowledge (or prior experience) is available. This paper develops a modeling framework integrating knowledge graph (KG) and probability-based recommendation algorithms to produce SWW treatment policies without depending on conventional physically based mathematical models and their associated quantitative data. The model is applied to the shale regions in China, including Chongqing city, the provinces of Sichuan, Yunnan, Guizhou, Shaanxi, and Inner Mongolia Autonomous Region, and utilizes the Monte-Carlo (MC) technique to assess the stability of the output policies. Results show that chemical precipitation, ultrafiltration, membrane bioreactor, and electrodialysis in Knowledge Graph Convolutional Networks (KGCN), as well as ultrafiltration, electrocoagulation, reverse osmosis, and electrocatalytic oxidation in RippleNet, are all preferentially recommended in one scenario, with the MC-based probabilities all higher than 0.86, indicating the stability of recommendation results and robustness of the model. Future studies would focus on improving the current KG and algorithms and offering explanatory mechanisms behind policy recommendations.

OpenPVSignal Knowledge Graph: Pharmacovigilance Signal Reports in a Computationally Exploitable FAIR Representation

INTRODUCTION

Pharmacovigilance signal report (PVSR) documents contain valuable condensed information published by drug monitoring organizations, typically in a free-text format. They provide initial insights into potential links between drugs and harmful effects. Still, their unstructured format prevents this valuable information from being integrated into data-processing pipelines (e.g., to support either the investigation of drug safety signals or decision-making in the clinical context).

OBJECTIVE

OpenPVSignal is a data model designed specifically to publish PVSRs via a computationally exploitable format, compliant with the FAIR (Findable, Accessible, Interoperable, Reusable) principles to promote ease of access and reusability of these valuable data.

METHODS

This paper outlines the procedure for converting pharmacovigilance signals published by the World Health Organization Uppsala Monitoring Centre (WHO-UMC) into the OpenPVSignal data model, resulting in a Knowledge Graph (KG). It details each step of the process, including the technical validation by KG engineers and the qualitative verification by medical and pharmacovigilance experts, leading to the finalized KG.

RESULTS

A total of 101 PVSRs from 2011 to 2019 were incorporated into the openly available KG.

CONCLUSION

The presented KG could be useful in various data-processing pipelines, including systems that support drug safety activities.

Differences in biomarker levels and proteomic survival prediction across two COVID-19 cohorts with distinct treatments

Prognostic biomarkers have been widely studied in COVID-19, but their levels may be influenced by treatment strategies. This study examined plasma biomarkers and proteomic survival prediction in two unvaccinated hospitalized COVID-19 cohorts receiving different treatments. In a derivation cohort (n = 126) from early 2020, we performed plasma proteomic profiling and evaluated innate and complement system immune markers. A proteomic model based on differentially expressed proteins predicted 30-day mortality with an area under the curve (AUC) of 0.81. The model was tested in a validation cohort (n = 80) from late 2020, where patients received remdesivir and dexamethasone, and performed with an AUC of 0.75. Biomarker levels varied considerably between cohorts, sometimes in opposite directions, highlighting the impact of treatment regimens on biomarker expression. These findings underscore the need to account for treatment effects when developing prognostic models, as treatment differences may limit their generalizability across populations.

MedKG: enabling drug discovery through a unified biomedical knowledge graph

Biomedical knowledge graphs have emerged as powerful tools for drug discovery, but existing platforms often suffer from outdated information, limited accessibility, and insufficient integration of complex data. This study presents MedKG, a comprehensive and continuously updated knowledge graph designed to address these challenges in precision medicine and drug discovery. MedKG integrates data from 35 authoritative sources, encompassing 34 node types and 79 relationships. A Continuous Integration/Continuous Update pipeline ensures MedKG remains current, addressing a critical limitation of static knowledge bases. The integration of molecular embeddings enhances semantic analysis capabilities, bridging the gap between chemical structures and biological entities. To demonstrate MedKG's utility, a novel hybrid Relational Graph Convolutional Network for disease-drug link prediction, MedLINK was developed and used in case studies on clinical trial data for disease drug link prediction. Furthermore, a web-based application with user-friendly APIs and visualization tools was built, making MedKG accessible to both technical and non-technical users, which is freely available at http://pitools.niper.ac.in/medkg/.

A clinical knowledge graph-based framework to prioritize candidate genes for facilitating diagnosis of Mendelian diseases and rare genetic conditions

BACKGROUND

Diagnosing Mendelian and rare genetic conditions requires identifying phenotype-associated genetic findings and prioritizing likely disease-causing genes. This task is labor-intensive for molecular and clinical geneticists, who must review extensive literature and databases to link patient phenotypes with causal genotypes. The challenge is further complicated by the large number of genetic variants detected through next-generation sequencing, which impacts both diagnosis timelines and patient care strategies. To address this, in silico methods that prioritize causal genes based on patient-derived phenotypes offer an effective solution, reducing the time involved in diagnostic case reviews and enhancing the efficiency of clinical diagnosis.

RESULTS

We developed the phenotype prioritization and analysis for rare diseases (PPAR) to rank genes based on human phenotype ontology (HPO) terms, with the specific goal of aiding the interpretation of genetic testing for Mendelian and rare diseases. PPAR leverages embeddings from a knowledge graph and incorporates knowledge from connections between genes, HPO terms, and gene ontology annotations. When applied on a clinical rare disease cohort and the publicly available deciphering developmental disorders (DDD) dataset. PPAR ranked the causal gene in the top 10 for 27% of cases in the clinical cohort and for 85% of cases in the DDD dataset, outperforming other established HPO-based methods.

CONCLUSION

Our findings demonstrate that PPAR, a method developed from the clinical knowledge graph, effectively ranks causal genes based on patient-derived HPO terms in rare and Mendelian disease contexts. PPAR has shown superior performance compared to other well-established HPO-only methods and provides an efficient, accessible solution for clinical geneticists. The Python-based tool is publicly available at https://github.com/dimi-lab/PPAR , offering a user-friendly platform for gene prioritization.

Multiomics Identifies Potential Molecular Profiles Associated With Outcomes After BRAF-Targeted Therapy in Patients With BRAF V600E-Mutated Advanced Solid Tumors

PURPOSE

It is a clinical challenge to select patients for BRAF-targeted therapy because of the lack of predictive biomarkers besides the BRAF V600E mutation. By analyzing the genome, transcriptome, and proteome, this study investigated the association between baseline molecular alterations and outcomes of BRAF-targeted therapy.

PATIENTS AND METHODS

Fresh tumor tissue from patients enrolled in the Copenhagen Prospective Personalized Oncology study was collected and underwent comprehensive molecular profiling.

RESULTS

TP53 comutations were most frequently detected. Patients with a TP53 wild-type tumor had a significantly longer median progression-free survival than those with TP53 comutations (hazard ratio, 2.8 [95% CI, 1.13 to 7.08]; P = .02). RNAseq revealed a distinct gene expression signature for patients with long-term disease control (LDC), including hallmarks of cell cycle arrest and proliferation in the p53 pathway. The protein analysis demonstrated that ubiquitin-conjugating enzyme EK2 was significantly downregulated in patients with LDC.

CONCLUSION

Using a multiomic approach, we identified molecular alterations associated with treatment outcomes. The potential of analyzing multiomic data is promising and should be prioritized in translational cancer research to uncover the full potential within precision oncology.

Mass spectrometry-based analysis of eccrine sweat supports predictive, preventive and personalised medicine in a cohort of breast cancer patients in Austria

Objective

Metabolomics measurements of eccrine sweat may provide novel and relevant biomedical information to support predictive, preventive and personalised medicine (3PM). However, only limited data is available regarding metabolic alterations accompanying chemotherapy of breast cancer patients related to residual cancer burden (RCB) or therapy response. Here, we have applied Metabo-Tip, a non-invasive metabolomics assay based on the analysis of eccrine sweat from the fingertips, to investigate the feasibility of such an approach, especially with respect to drug monitoring, assessing lifestyle parameters and stratification of breast cancer patients.

Methods

Eccrine sweat samples were collected from breast cancer patients (n = 9) during the first cycle of neoadjuvant chemotherapy at four time points in this proof-of-concept study at a Tertiary University Hospital. Metabolites in eccrine sweat were analysed using mass spectrometry. Blood plasma samples from the same timepoints were also collected and analysed using a validated targeted metabolomics kit, in addition to proteomics and fatty acids/oxylipin analysis.

Results

A total of 247 exogenous small molecules and endogenous metabolites were identified in eccrine sweat of the breast cancer patients. Cyclophosphamide and ondansetron were successfully detected and monitored in eccrine sweat of individual patients and accurately reflected the administration schedule. The non-essential amino acids asparagine, serine and proline, as well as ornithine were significantly regulated in eccrine sweat and blood plasma over the therapy cycle. However, their distinct time-dependent profiles indicated compartment-specific distributions. Indeed, the metabolite composition of eccrine sweat seems to largely resemble the composition of the interstitial fluid. Plasma proteins and fatty acids/oxylipins were not affected by the first treatment cycle. Individual smoking habit was revealed by the simultaneous detection of nicotine and its primary metabolite cotinine in eccrine sweat. Stratification according to RCB revealed pronounced differences in the metabolic composition of eccrine sweat in these patients at baseline, e.g., essential amino acids, possibly due to the systemic contribution of breast cancer and its impact on metabolic turnover.

Conclusion

Mass spectrometry-based analysis of metabolites from eccrine sweat of breast cancer patients successfully qualified lifestyle parameters for risk assessment and allowed us to monitor drug treatment and systemic response to therapy. Moreover, eccrine sweat revealed a potentially predictive metabolic pattern stratifying patients by the extent of the metabolic activity of breast cancer tissue at baseline. Eccrine sweat is derived from the otherwise hardly accessible interstitial fluid and, thus, opens up a new dimension for biomonitoring of breast cancer in secondary and tertiary care. The simple sample collection without the need for trained personnel could also enable decentralised long-term biomonitoring to assess stable disease or disease progression. Eccrine sweat analysis may indeed significantly advance 3PM for the benefit of breast cancer patients.

Supplementary Information

The online version contains supplementary material available at 10.1007/s13167-025-00396-6.

Mass-spectrometry-based proteomics: from single cells to clinical applications

Mass-spectrometry (MS)-based proteomics has evolved into a powerful tool for comprehensively analysing biological systems. Recent technological advances have markedly increased sensitivity, enabling single-cell proteomics and spatial profiling of tissues. Simultaneously, improvements in throughput and robustness are facilitating clinical applications. In this Review, we present the latest developments in proteomics technology, including novel sample-preparation methods, advanced instrumentation and innovative data-acquisition strategies. We explore how these advances drive progress in key areas such as protein-protein interactions, post-translational modifications and structural proteomics. Integrating artificial intelligence into the proteomics workflow accelerates data analysis and biological interpretation. We discuss the application of proteomics to single-cell analysis and spatial profiling, which can provide unprecedented insights into cellular heterogeneity and tissue architecture. Finally, we examine the transition of proteomics from basic research to clinical practice, including biomarker discovery in body fluids and the promise and challenges of implementing proteomics-based diagnostics. This Review provides a broad and high-level overview of the current state of proteomics and its potential to revolutionize our understanding of biology and transform medical practice.

Connecting electronic health records to a biomedical knowledge graph to link clinical phenotypes and molecular endotypes in atopic dermatitis

Precision medicine is defined by the U.S. Food & Drug Administration as "an innovative approach to tailoring disease prevention and treatment that considers differences in people's genes, environments, and lifestyles". To succeed in providing personalized medicine to patients, it will be necessary to integrate medical, biological and molecular data in order to identify all complex disease subtypes and understand their pathobiological mechanism. Since biomedical knowledge graphs (BKGs) are limited to the integration of prior knowledge data and do not integrate real-world data (RWD) that would allow for the incorporation of patient level information, we propose a first step towards using RWD, BKGs and graph machine learning (ML) to enable a fully integrated precision medicine strategy. In this study, we established a link between RWD and a BKG. Our methodology introduced a novel patient representation using graph ML applied to the BKG. This approach facilitated the interpretation and extension of ML findings, particularly in disease subtype identification with molecular data contained in the BKG. We applied our innovative methodology to deepen our understanding of atopic dermatitis, a condition with a complex underlying pathophysiological mechanism. Through our analysis, we identified seven subgroups of patients each characterized by clinical and genomic characteristics.

Rapid diagnosis and recurrence prediction of choledocholithiasis disease using raw bile with machine learning assisted SERS

Surface-enhanced Raman spectroscopy (SERS) analysis based on body fluids has been widely applied in disease diagnose. Choledocholithiasis is a widespread and often recurrent digestive system disease, with limited data on factors predicting its formation and reappearance. Bile contains many components that could provide valuable diagnostic information; however, the current diagnosis of biliary disease by SERS focuses on detecting specific component in the bile, overlooking the complex interplay and correlations among multiple factors that could be crucial for accurate diagnosis. This work directly obtained multi-component SERS spectral information of raw bile from 46 patients. Characteristic information was extracted from bile SERS spectra using Principal Component Analysis (PCA), revealing variations in the content of characteristic components associated with different choledocholithiasis types and their recurrence frequency. Pearson correlation analysis was also introduced to reveal the interactions of primary active substances pertinent to choledocholithiasis diagnosis. The efficacy of PCA and Support Vector Machine (SVM) models in classifying stone types, presented an accuracy of 99.2 %. Furthermore, the interaction patterns among SERS characteristic components in choledocholithiasis recurrence frequency were revealed, and with the support of SVM, the prediction for different recurrence rates reached an accuracy of 95.2 %. Overall, this work demonstrates that integrating SERS with machine learning can support disease diagnosis and the interpretation of correlations among multiple components, facilitating elucidating the disease mechanisms.

Redundant pathways for removal of defective RNA polymerase II complexes at a promoter-proximal pause checkpoint

The biological purpose of Integrator and RNA polymerase II (RNAPII) promoter-proximal pausing remains uncertain. Here, we show that loss of INTS6 in human cells results in increased interaction of RNAPII with proteins that can mediate its dissociation from the DNA template, including the CRL3ARMC5 E3 ligase, which ubiquitylates CTD serine5-phosphorylated RPB1 for degradation. ARMC5-dependent RNAPII ubiquitylation is activated by defects in factors acting at the promoter-proximal pause, including Integrator, DSIF, and capping enzyme. This ARMC5 checkpoint normally curtails a sizeable fraction of RNAPII transcription, and ARMC5 knockout cells produce more uncapped transcripts. When both the Integrator and CRL3ARMC5 turnover mechanisms are compromised, cell growth ceases and RNAPII with high pausing propensity disperses from the promoter-proximal pause site into the gene body. These data support a model in which CRL3ARMC5 functions alongside Integrator in a checkpoint mechanism that removes faulty RNAPII complexes at promoter-proximal pause sites to safeguard transcription integrity.

Unified Clinical Vocabulary Embeddings for Advancing Precision Medicine

Integrating clinical knowledge into AI remains challenging despite numerous medical guidelines and vocabularies. Medical codes, central to healthcare systems, often reflect operational patterns shaped by geographic factors, national policies, insurance frameworks, and physician practices rather than the precise representation of clinical knowledge. This disconnect hampers AI in representing clinical relationships, raising concerns about bias, transparency, and generalizability. Here, we developed a resource of 67,124 clinical vocabulary embeddings derived from a clinical knowledge graph tailored to electronic health record vocabularies, spanning over 1.3 million edges. Using graph transformer neural networks, we generated clinical vocabulary embeddings that provide a new representation of clinical knowledge by unifying seven medical vocabularies. These embeddings were validated through a phenotype risk score analysis involving 4.57 million patients from Clalit Healthcare Services, effectively stratifying individuals based on survival outcomes. Inter-institutional panels of clinicians evaluated the embeddings for alignment with clinical knowledge across 90 diseases and 3,000 clinical codes, confirming their robustness and transferability. This resource addresses gaps in integrating clinical vocabularies into AI models and training datasets, paving the way for knowledge-grounded population and patient-level models.

The use of 4D data-independent acquisition-based proteomic analysis and machine learning to reveal potential biomarkers for stress levels

Research suggests that individuals who experience prolonged exposure to stress may be at higher risk for developing psychological stress disorders. Currently, psychological stress is primarily evaluated by professional physicians using rating scales, which may be prone to subjective biases and limitations of the scales. Therefore, it is imperative to explore more objective, accurate, and efficient biomarkers for evaluating the level of psychological stress in an individual. In this study, we utilized 4D data-independent acquisition (4D-DIA) proteomics for quantitative protein analysis, and then employed support vector machine (SVM) combined with SHAP interpretation algorithm to identify potential biomarkers for psychological stress levels. Biomarkers validation was subsequently achieved through machine learning classification and a substantial amount of a priori knowledge derived from the knowledge graph. We performed cross-validation of the biomarkers using two batches of data, and the results showed that the combination of Glyceraldehyde-3-phosphate dehydrogenase and Fibronectin yielded an average area under the curve (AUC) of 92%, an average accuracy of 86%, an average F1 score of 79%, and an average sensitivity of 83%. Therefore, this combination may represent a potential approach for detecting stress levels to prevent psychological stress disorders.

Comprehensive applications of the artificial intelligence technology in new drug research and development

Purpose

Target-based strategy is a prevalent means of drug research and development (R&D), since targets provide effector molecules of drug action and offer the foundation of pharmacological investigation. Recently, the artificial intelligence (AI) technology has been utilized in various stages of drug R&D, where AI-assisted experimental methods show higher efficiency than sole experimental ones. It is a critical need to give a comprehensive review of AI applications in drug R &D for biopharmaceutical field.

Methods

Relevant literatures about AI-assisted drug R&D were collected from the public databases (Including Google Scholar, Web of Science, PubMed, IEEE Xplore Digital Library, Springer, and ScienceDirect) through a keyword searching strategy with the following terms [("Artificial Intelligence" OR "Knowledge Graph" OR "Machine Learning") AND ("Drug Target Identification" OR "New Drug Development")].

Results

In this review, we first introduced common strategies and novel trends of drug R&D, followed by characteristic description of AI algorithms widely used in drug R&D. Subsequently, we depicted detailed applications of AI algorithms in target identification, lead compound identification and optimization, drug repurposing, and drug analytical platform construction. Finally, we discussed the challenges and prospects of AI-assisted methods for drug discovery.

Conclusion

Collectively, this review provides comprehensive overview of AI applications in drug R&D and presents future perspectives for biopharmaceutical field, which may promote the development of drug industry.

KSDKG: construction and application of knowledge graph for kidney stone disease based on biomedical literature and public databases

Purpose

Kidney stone disease (KSD) is a common urological disorder with an increasing incidence worldwide. The extensive knowledge about KSD is dispersed across multiple databases, challenging the visualization and representation of its hierarchy and connections. This paper aims at constructing a disease-specific knowledge graph for KSD to enhance the effective utilization of knowledge by medical professionals and promote clinical research and discovery.

Methods

Text parsing and semantic analysis were conducted on literature related to KSD from PubMed, with concept annotation based on biomedical ontology being utilized to generate semantic data in RDF format. Moreover, public databases were integrated to construct a large-scale knowledge graph for KSD. Additionally, case studies were carried out to demonstrate the practical utility of the developed knowledge graph.

Results

We proposed and implemented a Kidney Stone Disease Knowledge Graph (KSDKG), covering more than 90 million triples. This graph comprised semantic data extracted from 29,174 articles, integrating available data from UMLS, SNOMED CT, MeSH, DrugBank and Microbe-Disease Knowledge Graph. Through the application of three cases, we retrieved and discovered information on microbes, drugs and diseases associated with KSD. The results illustrated that the KSDKG can integrate diverse medical knowledge and provide new clinical insights for identifying the underlying mechanisms of KSD.

Conclusion

The KSDKG efficiently utilizes knowledge graph to reveal hidden knowledge associations, facilitating semantic search and response. As a blueprint for developing disease-specific knowledge graphs, it offers valuable contributions to medical research.

Sleep deprivation leads to non-adaptive alterations in sleep microarchitecture and amyloid-β accumulation in a murine Alzheimer model

Impaired sleep is a common aspect of aging and often precedes the onset of Alzheimer's disease. Here, we compare the effects of sleep deprivation in young wild-type mice and their APP/PS1 littermates, a murine model of Alzheimer's disease. After 7 h of sleep deprivation, both genotypes exhibit an increase in EEG slow-wave activity. However, only the wild-type mice demonstrate an increase in the power of infraslow norepinephrine oscillations, which are characteristic of healthy non-rapid eye movement sleep. Notably, the APP/PS1 mice fail to enhance norepinephrine oscillations 24 h after sleep deprivation, coinciding with an accumulation of cerebral amyloid-β protein. Proteome analysis of cerebrospinal fluid and extracellular fluid further supports these findings by showing altered protein clearance in APP/PS1 mice. We propose that the suppression of infraslow norepinephrine oscillations following sleep deprivation contributes to increased vulnerability to sleep loss and heightens the risk of developing amyloid pathology in early stages of Alzheimer's disease.

BioGSF: a graph-driven semantic feature integration framework for biomedical relation extraction

The automatic and accurate extraction of diverse biomedical relations from literature constitutes the core elements of medical knowledge graphs, which are indispensable for healthcare artificial intelligence. Currently, fine-tuning through stacking various neural networks on pre-trained language models (PLMs) represents a common framework for end-to-end resolution of the biomedical relation extraction (RE) problem. Nevertheless, sequence-based PLMs, to a certain extent, fail to fully exploit the connections between semantics and the topological features formed by these connections. In this study, we presented a graph-driven framework named BioGSF for RE from the literature by integrating shortest dependency paths (SDP) with entity-pair graph through the employment of the graph neural network model. Initially, we leveraged dependency relationships to obtain the SDP between entities and incorporated this information into the entity-pair graph. Subsequently, the graph attention network was utilized to acquire the topological information of the entity-pair graph. Ultimately, the obtained topological information was combined with the semantic features of the contextual information for relation classification. Our method was evaluated on two distinct datasets, namely S4 and BioRED. The outcomes reveal that BioGSF not only attains the superior performance among previous models with a micro-F1 score of 96.68% (S4) and 96.03% (BioRED), but also demands the shortest running times. BioGSF emerges as an efficient framework for biomedical RE.

An improved cancer diagnosis algorithm for protein mass spectrometry based on PCA and a one-dimensional neural network combining ResNet and SENet

Cancer is one of the most serious health problems worldwide. Because cancer has no specific symptoms in its early stages, it is often not diagnosed until it is in advanced stages, reducing the likelihood of successful treatment. Therefore, early diagnosis of cancer is a formidable challenge. Mass spectrometry-based proteomics offers a robust technical foundation for cancer diagnosis. However, mass spectrometry data are characterized by high dimensionality, large data volume, and noise interference, which can lead to diagnostic errors in clinical applications. To address this challenge, an improved algorithm combining principal component analysis (PCA) with a convolutional neural network (CNN) algorithm (denoted as PCA-1DSE-ResCNN) was proposed to assist in analyzing high-dimensional mass spectral data. The algorithm initially reduced the dimensionality of the data through the PCA technique. Subsequently, the convolutional neural network algorithm (1DSE-ResCNN) integrating residual blocks and squeeze-and-excitation blocks was used as a classifier. This approach can not only alleviate the issues of overfitting and gradient vanishing caused by deep network layers but also reduce redundant information, enabling the algorithm to effectively learn high-dimensional data features and deal with nonlinear relationships. To validate the effectiveness of the algorithm, the high-dimensional ovarian cancer mass spectrometry dataset was selected as an example to examine its application performance in early diagnosis of ovarian cancer. The experimental results demonstrated that the PCA-1DSE-ResCNN algorithm outperforms other methods in terms of accuracy, specificity, and sensitivity on three high-dimensional ovarian cancer datasets. This study will contribute to the rapid diagnosis and early detection of cancer.

Generating Biomedical Knowledge Graphs from Knowledge Bases, Registries, and Multiomic Data

As large clinical and multiomics datasets and knowledge resources accumulate, they need to be transformed into computable and actionable information to support automated reasoning. These datasets range from laboratory experiment results to electronic health records (EHRs). Barriers to accessibility and sharing of such datasets include diversity of content, size and privacy. Effective transformation of data into information requires harmonization of stakeholder goals, implementation, enforcement of standards regarding quality and completeness, and availability of resources for maintenance and updates. Systems such as the Biomedical Data Translator leverage knowledge graphs (KGs), structured and machine learning readable knowledge representation, to encode knowledge extracted through inference. We focus here on the transformation of data from multiomics datasets and EHRs into compact knowledge, represented in a KG data structure. We demonstrate this data transformation in the context of the Translator ecosystem, including clinical trials, drug approvals, cancer, wellness, and EHR data. These transformations preserve individual privacy. We provide access to the five resulting KGs through the Translator framework. We show examples of biomedical research questions supported by our KGs, and discuss issues arising from extracting biomedical knowledge from multiomics data.

Multisource representation learning for pediatric knowledge extraction from electronic health records

Electronic Health Record (EHR) systems are particularly valuable in pediatrics due to high barriers in clinical studies, but pediatric EHR data often suffer from low content density. Existing EHR code embeddings tailored for the general patient population fail to address the unique needs of pediatric patients. To bridge this gap, we introduce a transfer learning approach, MUltisource Graph Synthesis (MUGS), aimed at accurate knowledge extraction and relation detection in pediatric contexts. MUGS integrates graphical data from both pediatric and general EHR systems, along with hierarchical medical ontologies, to create embeddings that adaptively capture both the homogeneity and heterogeneity between hospital systems. These embeddings enable refined EHR feature engineering and nuanced patient profiling, proving particularly effective in identifying pediatric patients similar to specific profiles, with a focus on pulmonary hypertension (PH). MUGS embeddings, resistant to negative transfer, outperform other benchmark methods in multiple applications, advancing evidence-based pediatric research.

Predicting drug outcome of population via clinical knowledge graph

Optimal treatments depend on numerous factors such as drug chemical properties, disease biology, and patient characteristics to which the treatment is applied. To realize the promise of AI in healthcare, there is a need for designing systems that can capture patient heterogeneity and relevant biomedical knowledge. Here we present PlaNet, a geometric deep learning framework that reasons over population variability, disease biology, and drug chemistry by representing knowledge in the form of a massive clinical knowledge graph that can be enhanced by language models. Our framework is applicable to any sub-population, any drug as well drug combinations, any disease, and a wide range of pharmacological tasks. We apply the PlaNet framework to reason about outcomes of clinical trials: PlaNet predicts drug efficacy and adverse events, even for experimental drugs and their combinations that have never been seen by the model. Furthermore, PlaNet can estimate the effect of changing population on trial outcomes with direct implications for patient stratification in clinical trials. PlaNet takes fundamental steps towards AI-guided clinical trials design, offering valuable guidance for realizing the vision of precision medicine using AI.

Improving Molecule Generation and Drug Discovery with a Knowledge-enhanced Generative Model

Recent advancements in generative models have established state-of-the-art benchmarks in the generation of molecules and novel drug candidates. Despite these successes, a significant gap persists between generative models and the utilization of extensive biomedical knowledge, often systematized within knowledge graphs, whose potential to inform and enhance generative processes has not been realized. In this paper, we present a novel approach that bridges this divide by developing a framework for knowledge-enhanced generative models called KARL. We develop a scalable methodology to extend the functionality of knowledge graphs while preserving semantic integrity, and incorporate this contextual information into a generative framework to guide a diffusion-based model. The integration of knowledge graph embeddings with our generative model furnishes a robust mechanism for producing novel drug candidates possessing specific characteristics while ensuring validity and synthesizability. KARL outperforms state-of-the-art generative models on both unconditional and targeted generation tasks.

Myosin ATPase inhibition fails to rescue the metabolically dysregulated proteome of nebulin-deficient muscle

Nemaline myopathy (NM) is a genetic muscle disease, primarily caused by mutations in the NEB gene (NEB-NM) and with muscle myosin dysfunction as a major molecular pathogenic mechanism. Recently, we have observed that the myosin biochemical super-relaxed state was significantly impaired in NEB-NM, inducing an aberrant increase in ATP consumption and remodelling of the energy proteome in diseased muscle fibres. Because the small-molecule Mavacamten is known to promote the myosin super-relaxed state and reduce the ATP demand, we tested its potency in the context of NEB-NM. We first conducted in vitro experiments in isolated single myofibres from patients and found that Mavacamten successfully reversed the myosin ATP overconsumption. Following this, we assessed its short-term in vivo effects using the conditional nebulin knockout (cNeb KO) mouse model and subsequently performing global proteomics profiling in dissected soleus myofibres. After a 4 week treatment period, we observed a remodelling of a large number of proteins in both cNeb KO mice and their wild-type siblings. Nevertheless, these changes were not related to the energy proteome, indicating that short-term Mavacamten treatment is not sufficient to properly counterbalance the metabolically dysregulated proteome of cNeb KO mice. Taken together, our findings emphasize Mavacamten potency in vitro but challenge its short-term efficacy in vivo. KEY POINTS: No cure exists for nemaline myopathy, a type of genetic skeletal muscle disease mainly derived from mutations in genes encoding myofilament proteins. Applying Mavacamten, a small molecule directly targeting the myofilaments, to isolated membrane-permeabilized muscle fibres from human patients restored myosin energetic disturbances. Treating a mouse model of nemaline myopathy in vivo with Mavacamten for 4 weeks, remodelled the skeletal muscle fibre proteome without any noticeable effects on energetic proteins. Short-term Mavacamten treatment may not be sufficient to reverse the muscle phenotype in nemaline myopathy.

Deficient brain GABA metabolism leads to widespread impairments of astrocyte and oligodendrocyte function

The neurometabolic disorder succinic semialdehyde dehydrogenase (SSADH) deficiency leads to great neurochemical imbalances and severe neurological manifestations. The cause of the disease is loss of function of the enzyme SSADH, leading to impaired metabolism of the principal inhibitory neurotransmitter GABA. Despite the known identity of the enzymatic deficit, the underlying pathology of SSADH deficiency remains unclear. To uncover new mechanisms of the disease, we performed an untargeted integrative analysis of cerebral protein expression, functional metabolism, and lipid composition in a genetic mouse model of SSADH deficiency (ALDH5A1 knockout mice). Our proteomic analysis revealed a clear regional vulnerability, as protein alterations primarily manifested in the hippocampus and cerebral cortex of the ALDH5A1 knockout mice. These regions displayed aberrant expression of proteins linked to amino acid homeostasis, mitochondria, glial function, and myelination. Stable isotope tracing in acutely isolated brain slices demonstrated an overall maintained oxidative metabolism of glucose, but a selective decrease in astrocyte metabolic activity in the cerebral cortex of ALDH5A1 knockout mice. In contrast, an elevated capacity of oxidative glutamine metabolism was observed in the ALDH5A1 knockout brain, which may serve as a neuronal compensation of impaired astrocyte glutamine provision. In addition to reduced expression of critical oligodendrocyte proteins, a severe depletion of myelin-enriched sphingolipids was found in the brains of ALDH5A1 knockout mice, suggesting degeneration of myelin. Altogether, our study highlights that impaired astrocyte and oligodendrocyte function is intimately linked to SSADH deficiency pathology, suggesting that selective targeting of glial cells may hold therapeutic potential in this disease.

Unlocking the secrets of NPSLE: the role of dendritic cell-secreted CCL2 in blood-brain barrier disruption

Background

This study employed RNA-seq technology and meta-analysis to unveil the molecular mechanisms of neuropsychiatric systemic lupus erythematosus (NPSLE) within the central nervous system.

Methods

Downloaded transcriptomic data on systemic lupus erythematosus (SLE) from the Gene Expression Omnibus (GEO) and analyzed differential genes in peripheral blood samples of NPSLE patients and healthy individuals. Employed WGCNA to identify key genes related to cognitive impairment and validated findings via RNA-seq. Conducted GO, KEGG, and GSEA analyses, and integrated PPI networks to explore gene regulatory mechanisms. Assessed gene impacts on dendritic cells and blood-brain barrier using RT-qPCR, ELISA, and in vitro models.

Results

Public databases and RNA-seq data have revealed a significant upregulation of CCL2 (C-C motif chemokine ligand 2) in the peripheral blood of both SLE and NPSLE patients, primarily secreted by mature dendritic cells. Furthermore, the secretion of CCL2 by mature dendritic cells may act through the RSAD2-ISG15 axis and is associated with the activation of the NLRs (Nod Like Receptor Signaling Pathway) signaling pathway in vascular endothelial cells. Subsequent in vitro cell experiments confirmed the high expression of CCL2 in peripheral blood dendritic cells of NPSLE patients, with its secretion being regulated by the RSAD2-ISG15 axis and inducing vascular endothelial cell pyroptosis through the activation of the NLRs signaling pathway. Clinical trial results ultimately confirmed that NPSLE patients exhibiting elevated CCL2 expression also experienced cognitive decline.

Conclusions

The secretion of CCL2 by dendritic cells induces pyroptosis in vascular endothelial cells, thereby promoting blood-brain barrier damage and triggering cognitive impairment in patients with systemic lupus erythematosus.

Graph databases in systems biology: a systematic review

Graph databases are becoming increasingly popular across scientific disciplines, being highly suitable for storing and connecting complex heterogeneous data. In systems biology, they are used as a backend solution for biological data repositories, ontologies, networks, pathways, and knowledge graph databases. In this review, we analyse all publications using or mentioning graph databases retrieved from PubMed and PubMed Central full-text search, focusing on the top 16 available graph databases, Publications are categorized according to their domain and application, focusing on pathway and network biology and relevant ontologies and tools. We detail different approaches and highlight the advantages of outstanding resources, such as UniProtKB, Disease Ontology, and Reactome, which provide graph-based solutions. We discuss ongoing efforts of the systems biology community to standardize and harmonize knowledge graph creation and the maintenance of integrated resources. Outlining prospects, including the use of graph databases as a way of communication between biological data repositories, we conclude that efficient design, querying, and maintenance of graph databases will be key for knowledge generation in systems biology and other research fields with heterogeneous data.

Knowledge Graphs for drug repurposing: a review of databases and methods

Drug repurposing has emerged as a effective and efficient strategy to identify new treatments for a variety of diseases. One of the most effective approaches for discovering potential new drug candidates involves the utilization of Knowledge Graphs (KGs). This review comprehensively explores some of the most prominent KGs, detailing their structure, data sources, and how they facilitate the repurposing of drugs. In addition to KGs, this paper delves into various artificial intelligence techniques that enhance the process of drug repurposing. These methods not only accelerate the identification of viable drug candidates but also improve the precision of predictions by leveraging complex datasets and advanced algorithms. Furthermore, the importance of explainability in drug repurposing is emphasized. Explainability methods are crucial as they provide insights into the reasoning behind AI-generated predictions, thereby increasing the trustworthiness and transparency of the repurposing process. We will discuss several techniques that can be employed to validate these predictions, ensuring that they are both reliable and understandable.

STAVER: a standardized benchmark dataset-based algorithm for effective variation reduction in large-scale DIA-MS data

Mass spectrometry (MS)-based proteomics has become instrumental in comprehensively investigating complex biological systems. Data-independent acquisition (DIA)-MS, utilizing hybrid spectral library search strategies, allows for the simultaneous quantification of thousands of proteins, showing promise in enhancing protein identification and quantification precision. However, low-quality profiles can considerably undermine quantitative precision, resulting in inaccurate protein quantification. To tackle this challenge, we introduced STAVER, a novel algorithm that leverages standardized benchmark datasets to reduce non-biological variation in large-scale DIA-MS analyses. By eliminating unwanted noise in MS signals, STAVER significantly improved protein quantification precision, especially in hybrid spectral library searches. Moreover, we validated STAVER's robustness and applicability across multiple large-scale DIA datasets, demonstrating significantly enhanced precision and reproducibility of protein quantification. STAVER offers an innovative and effective approach for enhancing the quality of large-scale DIA proteomic data, facilitating cross-platform and cross-laboratory comparative analyses. This advancement significantly enhances the consistency and reliability of findings in clinical research. The complete package is available at https://github.com/Ran485/STAVER.

µPhos: a scalable and sensitive platform for high-dimensional phosphoproteomics

Mass spectrometry has revolutionized cell signaling research by vastly simplifying the analysis of many thousands of phosphorylation sites in the human proteome. Defining the cellular response to perturbations is crucial for further illuminating the functionality of the phosphoproteome. Here we describe µPhos ('microPhos'), an accessible phosphoproteomics platform that permits phosphopeptide enrichment from 96-well cell culture and small tissue amounts in <8 h total processing time. By greatly minimizing transfer steps and liquid volumes, we demonstrate increased sensitivity, >90% selectivity, and excellent quantitative reproducibility. Employing highly sensitive trapped ion mobility mass spectrometry, we quantify ~17,000 Class I phosphosites in a human cancer cell line using 20 µg starting material, and confidently localize ~6200 phosphosites from 1 µg. This depth covers key signaling pathways, rendering sample-limited applications and perturbation experiments with hundreds of samples viable. We employ µPhos to study drug- and time-dependent response signatures in a leukemia cell line, and by quantifying 30,000 Class I phosphosites in the mouse brain we reveal distinct spatial kinase activities in subregions of the hippocampal formation.

The impact of short-term eucaloric low- and high-carbohydrate diets on liver triacylglycerol content in males with overweight and obesity: a randomized crossover study

BACKGROUND

Intrahepatic triacylglycerol (liver TG) content is associated with hepatic insulin resistance and dyslipidemia. Liver TG content can be modulated within days under hypocaloric conditions.

OBJECTIVES

We hypothesized that 4 d of eucaloric low-carbohydrate/high-fat (LC) intake would decrease liver TG content, whereas a high-carbohydrate/low-fat (HC) intake would increase liver TG content, and further that alterations in liver TG would be linked to dynamic changes in hepatic glucose and lipid metabolism.

METHODS

A randomized crossover trial in males with 4 d + 4 d of LC and HC, respectively, with ≥2 wk of washout. 1H-magnetic resonance spectroscopy (1H-MRS) was used to measure liver TG content, with metabolic testing before and after intake of an LC diet (11E% carbohydrate corresponding to 102 ± 12 {mean ± standard deviation [SD]) g/d, 70E% fat} and an HC diet (65E% carbohydrate corresponding to 537 ± 56 g/d, 16E% fat). Stable [6,6-2H2]-glucose and [1,1,2,3,3-D5]-glycerol tracer infusions combined with hyperinsulinemic-euglycemic clamps and indirect calorimetry were used to measure rates of hepatic glucose production and lipolysis, whole-body insulin sensitivity and substrate oxidation.

RESULTS

Eleven normoglycemic males with overweight or obesity (BMI 31.6 ± 3.7 kg/m2) completed both diets. The LC diet reduced liver TG content by 35.3% (95% confidence interval: -46.6, -24.1) from 4.9% [2.4-11.0] (median interquartile range) to 2.9% [1.4-6.9], whereas there was no change after the HC diet. After the LC diet, fasting whole-body fat oxidation and plasma beta-hydroxybutyrate concentration increased, whereas markers of de novo lipogenesis (DNL) diminished. Fasting plasma TG and insulin concentrations were lowered and the hepatic insulin sensitivity index increased after LC. Peripheral glucose disposal was unchanged.

CONCLUSIONS

Reduced carbohydrate and increased fat intake for 4 d induced a marked reduction in liver TG content and increased hepatic insulin sensitivity. Increased rates of fat oxidation and ketogenesis combined with lower rates of DNL are suggested to be responsible for lowering liver TG. This trial was registered at clinicaltrials.gov as NCT04581421.

Computational Approaches to Drug Repurposing: Methods, Challenges, and Opportunities

Drug repurposing refers to the inference of therapeutic relationships between a clinical indication and existing compounds. As an emerging paradigm in drug development, drug repurposing enables more efficient treatment of rare diseases, stratified patient populations, and urgent threats to public health. However, prioritizing well-suited drug candidates from among a nearly infinite number of repurposing options continues to represent a significant challenge in drug development. Over the past decade, advances in genomic profiling, database curation, and machine learning techniques have enabled more accurate identification of drug repurposing candidates for subsequent clinical evaluation. This review outlines the major methodologic classes that these approaches comprise, which rely on (a) protein structure, (b) genomic signatures, (c) biological networks, and (d) real-world clinical data. We propose that realizing the full impact of drug repurposing methodologies requires a multidisciplinary understanding of each method's advantages and limitations with respect to clinical practice.

Proteomic profile and predictive markers of outcome in patients with subarachnoid hemorrhage

BACKGROUND

The molecular mechanisms underlying development of posthemorrhagic hydrocephalus (PHH) following subarachnoid hemorrhage (SAH) remain incompletely understood. Consequently, treatment strategies tailored towards the individual patient remain limited. This study aimed to identify proteomic cerebrospinal fluid (CSF) biomarkers capable of predicting shunt dependency and functional outcome in patients with SAH in order to improve informed clinical decision making.

METHODS

Ventricular CSF samples were collected twice from 23 patients with SAH who required external ventricular drain (EVD) insertion (12 patients with successful EVD weaning, 11 patients in need of permanent CSF shunting due to development of PHH). The paired CSF samples were collected acutely after ictus and later upon EVD removal. Cisternal CSF samples were collected from 10 healthy control subjects undergoing vascular clipping of an unruptured aneurysm. All CSF samples were subjected to mass spectrometry-based proteomics analysis. Proteomic biomarkers were quantified using area under the curve (AUC) estimates from a receiver operating curve (ROC).

RESULTS

CSF from patients with SAH displayed a distinct proteomic profile in comparison to that of healthy control subjects. The CSF collected acutely after ictus from patients with SAH was moreover distinct from that collected weeks later but appeared similar in the weaned and shunted patient groups. Sixteen unique proteins were identified as potential predictors of shunt dependency, while three proteins were identified as potential predictors of functional outcome assessed six months after ictus with the modified Rankin Scale.

CONCLUSIONS

We here identified several potential proteomic biomarkers in CSF from patients with SAH capable of predicting (i) shunt dependency and thus development of PHH and (ii) the functional outcome assessed six months after ictus. These proteomic biomarkers may have the potential to aid clinical decision making by predicting shunt dependency and functional outcome following SAH.

An experimentally validated approach to automated biological evidence generation in drug discovery using knowledge graphs

Explaining predictions for drug repositioning with biological knowledge graphs is a challenging problem. Graph completion methods using symbolic reasoning predict drug treatments and associated rules to generate evidence representing the therapeutic basis of the drug. Yet the vast amounts of generated paths that are biologically irrelevant or not mechanistically meaningful within the context of disease biology can limit utility. We use a reinforcement learning based knowledge graph completion model combined with an automatic filtering approach that produces the most relevant rules and biological paths explaining the predicted drug's therapeutic connection to the disease. In this work we validate the approach against preclinical experimental data for Fragile X syndrome demonstrating strong correlation between automatically extracted paths and experimentally derived transcriptional changes of selected genes and pathways of drug predictions Sulindac and Ibudilast. Additionally, we show it reduces the number of generated paths in two case studies, 85% for Cystic fibrosis and 95% for Parkinson's disease.

Graph machine learning for integrated multi-omics analysis

Multi-omics experiments at bulk or single-cell resolution facilitate the discovery of hypothesis-generating biomarkers for predicting response to therapy, as well as aid in uncovering mechanistic insights into cellular and microenvironmental processes. Many methods for data integration have been developed for the identification of key elements that explain or predict disease risk or other biological outcomes. The heterogeneous graph representation of multi-omics data provides an advantage for discerning patterns suitable for predictive/exploratory analysis, thus permitting the modeling of complex relationships. Graph-based approaches-including graph neural networks-potentially offer a reliable methodological toolset that can provide a tangible alternative to scientists and clinicians that seek ideas and implementation strategies in the integrated analysis of their omics sets for biomedical research. Graph-based workflows continue to push the limits of the technological envelope, and this perspective provides a focused literature review of research articles in which graph machine learning is utilized for integrated multi-omics data analyses, with several examples that demonstrate the effectiveness of graph-based approaches.

Deep 3D histology powered by tissue clearing, omics and AI

To comprehensively understand tissue and organism physiology and pathophysiology, it is essential to create complete three-dimensional (3D) cellular maps. These maps require structural data, such as the 3D configuration and positioning of tissues and cells, and molecular data on the constitution of each cell, spanning from the DNA sequence to protein expression. While single-cell transcriptomics is illuminating the cellular and molecular diversity across species and tissues, the 3D spatial context of these molecular data is often overlooked. Here, I discuss emerging 3D tissue histology techniques that add the missing third spatial dimension to biomedical research. Through innovations in tissue-clearing chemistry, labeling and volumetric imaging that enhance 3D reconstructions and their synergy with molecular techniques, these technologies will provide detailed blueprints of entire organs or organisms at the cellular level. Machine learning, especially deep learning, will be essential for extracting meaningful insights from the vast data. Further development of integrated structural, molecular and computational methods will unlock the full potential of next-generation 3D histology.

Microbiome modeling: a beginner's guide

Microbiomes, comprised of diverse microbial species and viruses, play pivotal roles in human health, environmental processes, and biotechnological applications and interact with each other, their environment, and hosts via ecological interactions. Our understanding of microbiomes is still limited and hampered by their complexity. A concept improving this understanding is systems biology, which focuses on the holistic description of biological systems utilizing experimental and computational methods. An important set of such experimental methods are metaomics methods which analyze microbiomes and output lists of molecular features. These lists of data are integrated, interpreted, and compiled into computational microbiome models, to predict, optimize, and control microbiome behavior. There exists a gap in understanding between microbiologists and modelers/bioinformaticians, stemming from a lack of interdisciplinary knowledge. This knowledge gap hinders the establishment of computational models in microbiome analysis. This review aims to bridge this gap and is tailored for microbiologists, researchers new to microbiome modeling, and bioinformaticians. To achieve this goal, it provides an interdisciplinary overview of microbiome modeling, starting with fundamental knowledge of microbiomes, metaomics methods, common modeling formalisms, and how models facilitate microbiome control. It concludes with guidelines and repositories for modeling. Each section provides entry-level information, example applications, and important references, serving as a valuable resource for comprehending and navigating the complex landscape of microbiome research and modeling.

Proteomic profiling reveals diagnostic signatures and pathogenic insights in multisystem inflammatory syndrome in children

Multisystem inflammatory syndrome in children (MIS-C) is a severe disease that emerged during the COVID-19 pandemic. Although recognized as an immune-mediated condition, the pathogenesis remains unresolved. Furthermore, the absence of a diagnostic test can lead to delayed immunotherapy. Using state-of-the-art mass-spectrometry proteomics, assisted by artificial intelligence (AI), we aimed to identify a diagnostic signature for MIS-C and to gain insights into disease mechanisms. We identified a highly specific 4-protein diagnostic signature in children with MIS-C. Furthermore, we identified seven clusters that differed between MIS-C and controls, indicating an interplay between apolipoproteins, immune response proteins, coagulation factors, platelet function, and the complement cascade. These intricate protein patterns indicated MIS-C as an immunometabolic condition with global hypercoagulability. Our findings emphasize the potential of AI-assisted proteomics as a powerful and unbiased tool for assessing disease pathogenesis and suggesting avenues for future interventions and impact on pediatric disease trajectories through early diagnosis.

Clinical trial recommendations using Semantics-Based inductive inference and knowledge graph embeddings

OBJECTIVE

Designing a new clinical trial entails many decisions, such as defining a cohort and setting the study objectives to name a few, and therefore can benefit from recommendations based on exhaustive mining of past clinical trial records. This study proposes an approach based on knowledge graph embeddings and semantics-driven inductive inference for generating such recommendations.

METHOD

The proposed recommendation methodology is based on neural embeddings trained on first-of-its-kind knowledge graph constructed from clinical trials data. The methodology includes design of a knowledge graph for clinical trial data, evaluation of various knowledge graph embedding techniques for it, application of a novel inductive inference method using these embeddings, and generation of recommendations for clinical trial design. The study uses freely available data from clinicaltrials.gov and related sources.

RESULTS

The proposed approach for recommendations obtained relevance scores ranging from 70% to 83%. These scores were determined by evaluating the text similarity of recommended elements to actual elements used in clinical trials that are in progress. Furthermore, the most pertinent recommendations were consistently located towards the top of the list, indicating the effectiveness of our method.

CONCLUSION

Our study suggests that inductive inference using node semantics is a viable approach for generating recommendations using graphs neural embeddings, and that there is a potential for improvement in training graph embeddings using node semantics.

NBR1-mediated selective autophagy of ARF7 modulates root branching

Auxin dictates root architecture via the Auxin Response Factor (ARF) family of transcription factors, which control lateral root (LR) formation. In Arabidopsis, ARF7 regulates the specification of prebranch sites (PBS) generating LRs through gene expression oscillations and plays a pivotal role during LR initiation. Despite the importance of ARF7 in this process, there is a surprising lack of knowledge about how ARF7 turnover is regulated and how this impacts root architecture. Here, we show that ARF7 accumulates in autophagy mutants and is degraded through NBR1-dependent selective autophagy. We demonstrate that the previously reported rhythmic changes to ARF7 abundance in roots are modulated via autophagy and might occur in other tissues. In addition, we show that the level of co-localization between ARF7 and autophagy markers oscillates and can be modulated by auxin to trigger ARF7 turnover. Furthermore, we observe that autophagy impairment prevents ARF7 oscillation and reduces both PBS establishment and LR formation. In conclusion, we report a novel role for autophagy during development, namely by enacting auxin-induced selective degradation of ARF7 to optimize periodic root branching.

Remodeling of skeletal muscle myosin metabolic states in hibernating mammals

Hibernation is a period of metabolic suppression utilized by many small and large mammal species to survive during winter periods. As the underlying cellular and molecular mechanisms remain incompletely understood, our study aimed to determine whether skeletal muscle myosin and its metabolic efficiency undergo alterations during hibernation to optimize energy utilization. We isolated muscle fibers from small hibernators, Ictidomys tridecemlineatus and Eliomys quercinus and larger hibernators, Ursus arctos and Ursus americanus. We then conducted loaded Mant-ATP chase experiments alongside X-ray diffraction to measure resting myosin dynamics and its ATP demand. In parallel, we performed multiple proteomics analyses. Our results showed a preservation of myosin structure in U. arctos and U. americanus during hibernation, whilst in I. tridecemlineatus and E. quercinus, changes in myosin metabolic states during torpor unexpectedly led to higher levels in energy expenditure of type II, fast-twitch muscle fibers at ambient lab temperatures (20 °C). Upon repeating loaded Mant-ATP chase experiments at 8 °C (near the body temperature of torpid animals), we found that myosin ATP consumption in type II muscle fibers was reduced by 77-107% during torpor compared to active periods. Additionally, we observed Myh2 hyper-phosphorylation during torpor in I. tridecemilineatus, which was predicted to stabilize the myosin molecule. This may act as a potential molecular mechanism mitigating myosin-associated increases in skeletal muscle energy expenditure during periods of torpor in response to cold exposure. Altogether, we demonstrate that resting myosin is altered in hibernating mammals, contributing to significant changes to the ATP consumption of skeletal muscle. Additionally, we observe that it is further altered in response to cold exposure and highlight myosin as a potentially contributor to skeletal muscle non-shivering thermogenesis.

An open source knowledge graph ecosystem for the life sciences

Translational research requires data at multiple scales of biological organization. Advancements in sequencing and multi-omics technologies have increased the availability of these data, but researchers face significant integration challenges. Knowledge graphs (KGs) are used to model complex phenomena, and methods exist to construct them automatically. However, tackling complex biomedical integration problems requires flexibility in the way knowledge is modeled. Moreover, existing KG construction methods provide robust tooling at the cost of fixed or limited choices among knowledge representation models. PheKnowLator (Phenotype Knowledge Translator) is a semantic ecosystem for automating the FAIR (Findable, Accessible, Interoperable, and Reusable) construction of ontologically grounded KGs with fully customizable knowledge representation. The ecosystem includes KG construction resources (e.g., data preparation APIs), analysis tools (e.g., SPARQL endpoint resources and abstraction algorithms), and benchmarks (e.g., prebuilt KGs). We evaluated the ecosystem by systematically comparing it to existing open-source KG construction methods and by analyzing its computational performance when used to construct 12 different large-scale KGs. With flexible knowledge representation, PheKnowLator enables fully customizable KGs without compromising performance or usability.

Supramolecular Chiral Binding Affinity-Achieved Efficient Synergistic Cancer Therapy

Supramolecular chirality-mediated selective interaction among native assemblies is essential for precise disease diagnosis and treatment. Herein, to fully understand the supramolecular chiral binding affinity-achieved therapeutic efficiency, supramolecular chiral nanoparticles (WP5⊃D/L-Arg+DOX+ICG) with the chirality transfer from chiral arginine (D/L-Arg) to water-soluble pillar[5]arene (WP5) are developed through non-covalent interactions, in which an anticancer drug (DOX, doxorubicin hydrochloride) and a photothermal agent (ICG, indocyanine green) are successfully loaded. Interestingly, the WP5⊃D-Arg nanoparticles show 107 folds stronger binding capability toward phospholipid-composed liposomes compared with WP5⊃L-Arg. The enantioselective interaction further triggers the supramolecular chirality-specific drug accumulation in cancer cells. As a consequence, WP5⊃D-Arg+DOX+ICG exhibits extremely enhanced chemo-photothermal synergistic therapeutic efficacy (tumor inhibition rate of 99.4%) than that of WP5⊃L-Arg+DOX+ICG (tumor inhibition rate of 56.4%) under the same condition. This work reveals the breakthrough that supramolecular chiral assemblies can induce surprisingly large difference in cancer therapy, providing strong support for the significance of supramolecular chirality in bio-application.

Artificial Intelligence for Quantitative Modeling in Drug Discovery and Development: An Innovation and Quality Consortium Perspective on Use Cases and Best Practices

Recent breakthroughs in artificial intelligence (AI) and machine learning (ML) have ushered in a new era of possibilities across various scientific domains. One area where these advancements hold significant promise is model-informed drug discovery and development (MID3). To foster a wider adoption and acceptance of these advanced algorithms, the Innovation and Quality (IQ) Consortium initiated the AI/ML working group in 2021 with the aim of promoting their acceptance among the broader scientific community as well as by regulatory agencies. By drawing insights from workshops organized by the working group and attended by key stakeholders across the biopharma industry, academia, and regulatory agencies, this white paper provides a perspective from the IQ Consortium. The range of applications covered in this white paper encompass the following thematic topics: (i) AI/ML-enabled Analytics for Pharmacometrics and Quantitative Systems Pharmacology (QSP) Workflows; (ii) Explainable Artificial Intelligence and its Applications in Disease Progression Modeling; (iii) Natural Language Processing (NLP) in Quantitative Pharmacology Modeling; and (iv) AI/ML Utilization in Drug Discovery. Additionally, the paper offers a set of best practices to ensure an effective and responsible use of AI, including considering the context of use, explainability and generalizability of models, and having human-in-the-loop. We believe that embracing the transformative power of AI in quantitative modeling while adopting a set of good practices can unlock new opportunities for innovation, increase efficiency, and ultimately bring benefits to patients.

Application of artificial intelligence and machine learning in drug repurposing

The purpose of drug repurposing is to leverage previously approved drugs for a particular disease indication and apply them to another disease. It can be seen as a faster and more cost-effective approach to drug discovery and a powerful tool for achieving precision medicine. In addition, drug repurposing can be used to identify therapeutic candidates for rare diseases and phenotypic conditions with limited information on disease biology. Machine learning and artificial intelligence (AI) methodologies have enabled the construction of effective, data-driven repurposing pipelines by integrating and analyzing large-scale biomedical data. Recent technological advances, especially in heterogeneous network mining and natural language processing, have opened up exciting new opportunities and analytical strategies for drug repurposing. In this review, we first introduce the challenges in repurposing approaches and highlight some success stories, including those during the COVID-19 pandemic. Next, we review some existing computational frameworks in the literature, organized on the basis of the type of biomedical input data analyzed and the computational algorithms involved. In conclusion, we outline some exciting new directions that drug repurposing research may take, as pioneered by the generative AI revolution.

A Sample-Centric and Knowledge-Driven Computational Framework for Natural Products Drug Discovery

The ENPKG framework organizes large heterogeneous metabolomics data sets as a knowledge graph, offering exciting opportunities for drug discovery and chemodiversity characterization.

Predicting nuclear G-quadruplex RNA-binding proteins with roles in transcription and phase separation

RNA-binding proteins are central for many biological processes and their characterization has demonstrated a broad range of functions as well as a wide spectrum of target structures. RNA G-quadruplexes are important regulatory elements occurring in both coding and non-coding transcripts, yet our knowledge of their structure-based interactions is at present limited. Here, using theoretical predictions and experimental approaches, we show that many chromatin-binding proteins bind to RNA G-quadruplexes, and we classify them based on their RNA G-quadruplex-binding potential. Combining experimental identification of nuclear RNA G-quadruplex-binding proteins with computational approaches, we build a prediction tool that assigns probability score for a nuclear protein to bind RNA G-quadruplexes. We show that predicted G-quadruplex RNA-binding proteins exhibit a high degree of protein disorder and hydrophilicity and suggest involvement in both transcription and phase-separation into membrane-less organelles. Finally, we present the G4-Folded/UNfolded Nuclear Interaction Explorer System (G4-FUNNIES) for estimating RNA G4-binding propensities at http://service.tartaglialab.com/new_submission/G4FUNNIES .

Discovery of plasma proteins associated with ventricular fibrillation during first ST-elevation myocardial infarction via proteomics

AIMS

The underlying biological mechanisms of ventricular fibrillation (VF) during acute myocardial infarction are largely unknown. To our knowledge, this is the first proteomic study for this trait, with the aim to identify and characterize proteins that are associated with VF during first ST-elevation myocardial infarction (STEMI).

METHODS AND RESULTS

We included 230 participants from a Danish ongoing case-control study on patients with first STEMI with VF (case, n = 110) and without VF (control, n = 120) before guided catheter insertion for primary percutaneous coronary intervention. The plasma proteome was investigated using mass spectrometry-based proteomics on plasma samples collected within 24 h of symptom onset, and one patient was excluded in quality control. In 229 STEMI patients {72% men, median age 62 years [interquartile range (IQR): 54-70]}, a median of 257 proteins (IQR: 244-281) were quantified per patient. A total of 26 proteins were associated with VF; these proteins were involved in several biological processes including blood coagulation, haemostasis, and immunity. After correcting for multiple testing, two up-regulated proteins remained significantly associated with VF, actin beta-like 2 [ACTBL2, fold change (FC) 2.25, P < 0.001, q = 0.023], and coagulation factor XIII-A (F13A1, FC 1.48, P < 0.001, q = 0.023). None of the proteins were correlated with anterior infarct location.

CONCLUSION

Ventricular fibrillation due to first STEMI was significantly associated with two up-regulated proteins (ACTBL2 and F13A1), suggesting that they may represent novel underlying molecular VF mechanisms. Further research is needed to determine whether these proteins are predictive biomarkers or acute phase response proteins to VF during acute ischaemia.