Understanding the need for digital twins' data in patient advocacy and forecasting oncology.

Digital twins are made of a real-world component where data is measured and a virtual component where those measurements are used to parameterize computational models. There is growing interest in applying digital twins-based approaches to optimize personalized treatment plans and improve health outcomes. The integration of artificial intelligence is critical in this process, as it enables the development of sophisticated disease models that can accurately predict patient response to therapeutic interventions. There is a unique and equally important application of AI to the real-world component of a digital twin when it is applied to medical interventions. The patient can only be treated once, and therefore, we must turn to the experience and outcomes of previously treated patients for validation and optimization of the computational predictions. The physical component of a digital twins instead must utilize a compilation of available data from previously treated cancer patients whose characteristics (genetics, tumor type, lifestyle, etc.) closely parallel those of a newly diagnosed cancer patient for the purpose of predicting outcomes, stratifying treatment options, predicting responses to treatment and/or adverse events. These tasks include the development of robust data collection methods, ensuring data availability, creating precise and dependable models, and establishing ethical guidelines for the use and sharing of data. To successfully implement digital twin technology in clinical care, it is crucial to gather data that accurately reflects the variety of diseases and the diversity of the population.

Adaptively Integrative Association between Multivariate Phenotypes and Transcriptomic Data for Complex Diseases.

Phenotype-gene association studies can uncover disease mechanisms for translational research. Association with multiple phenotypes or clinical variables in complex diseases has the advantage of increasing statistical power and offering a holistic view. Existing multi-variate association methods mostly focus on SNP-based genetic associations. In this paper, we extend and evaluate two adaptive Fisher's methods, namely AFp and AFz, from the p-value combination perspective for phenotype-mRNA association analysis. The proposed method effectively aggregates heterogeneous phenotype-gene effects, allows association with different data types of phenotypes, and performs the selection of the associated phenotypes. Variability indices of the phenotype-gene effect selection are calculated by bootstrap analysis, and the resulting co-membership matrix identifies gene modules clustered by phenotype-gene effect. Extensive simulations demonstrate the superior performance of AFp compared to existing methods in terms of type I error control, statistical power and biological interpretation. Finally, the method is separately applied to three sets of transcriptomic and clinical datasets from lung disease, breast cancer, and brain aging and generates intriguing biological findings.

Genome-wide analyses identify novel risk loci for cluster headache in Han Chinese residing in Taiwan.

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.

Efficiently Predicting Vancomycin Resistance of Enterococcus Faecium From MALDI-TOF MS Spectra Using a Deep Learning-Based Approach.

Matrix-assisted laser desorption ionization time-of-flight (MALDI-TOF) mass spectrometry (MS) has recently become a useful analytical approach for microbial identification. The presence and absence of specific peaks on MS spectra are commonly used to identify the bacterial species and predict antibiotic-resistant strains. However, the conventional approach using few single peaks would result in insufficient prediction power without using complete information of whole MS spectra. In the past few years, machine learning algorithms have been successfully applied to analyze the MALDI-TOF MS peaks pattern for rapid strain typing. In this study, we developed a convolutional neural network (CNN) method to deal with the complete information of MALDI-TOF MS spectra for detecting Enterococcus faecium, which is one of the leading pathogens in the world. We developed a CNN model to rapidly and accurately predict vancomycin-resistant Enterococcus faecium (VREfm) samples from the whole mass spectra profiles of clinical samples. The CNN models demonstrated good classification performances with the average area under the receiver operating characteristic curve (AUROC) of 0.887 when using external validation data independently. Additionally, we employed the score-class activation mapping (CAM) method to identify the important features of our CNN models and found some discriminative signals that can substantially contribute to detecting the ion of resistance. This study not only utilized the complete information of MALTI-TOF MS data directly but also provided a practical means for rapid detection of VREfm using a deep learning algorithm.

Network hub-node prioritization of gene regulation with intra-network association.

BACKGROUND: To identify and prioritize the influential hub genes in a gene-set or biological pathway, most analyses rely on calculation of marginal effects or tests of statistical significance. These procedures may be inappropriate since hub nodes are common connection points and therefore may interact with other nodes more often than non-hub nodes do. Such dependence among gene nodes can be conjectured based on the topology of the pathway network or the correlation between them. RESULTS: Here we develop a pathway activity score incorporating the marginal (local) effects of gene nodes as well as intra-network affinity measures. This score summarizes the expression levels in a gene-set/pathway for each sample, with weights on local and network information, respectively. The score is next used to examine the impact of each node through a leave-one-out evaluation. To illustrate the procedure, two cancer studies, one involving RNA-Seq from breast cancer patients with high-grade ductal carcinoma in situ and one microarray expression data from ovarian cancer patients, are used to assess the performance of the procedure, and to compare with existing methods, both ones that do and do not take into consideration correlation and network information. The hub nodes identified by the proposed procedure in the two cancer studies are known influential genes; some have been included in standard treatments and some are currently considered in clinical trials for target therapy. The results from simulation studies show that when marginal effects are mild or weak, the proposed procedure can still identify causal nodes, whereas methods relying only on marginal effect size cannot. CONCLUSIONS: The NetworkHub procedure proposed in this research can effectively utilize the network information in combination with local effects derived from marker values, and provide a useful and complementary list of recommendations for prioritizing causal hubs.