Multiple Independent Datasets Research Articles

Abstract P1144: Single-cell RNA-Sequencing Of The Murine Cardiac Conduction System Reveals Transcriptional Regulators Associated With Conduction Development And Disease

Rhythmic beating of the heart is driven by a small subset of heart cells, altogether known as the cardiac conduction system (CCS). The CCS consists of unique components including the sinoatrial node (SAN), the atrioventricular node (AVN), the bundle of His, bundle branches and Purkinje fiber (PF) system. Due to obstacles inherent in studying the CCS, including small cell numbers, large cell-type heterogeneity and complex anatomy, our understanding of the transcriptional networks driving CCS development remains limited. Here, we harnessed single-cell RNA-sequencing (scRNAseq) and bioinformatic analyses to uncover putative DNA-binding transcription factors (TFs) as well as transcriptional co-regulators enriched within the entire developing murine CCS. Comparison of multiple independent prior datasets, spanning all major components of the CCS, revealed previously known TFs necessary for conduction development including, Islet 1 ( Isl1 ; avg log fold change 0.34, p= 1E-192) in the SAN; T-box transcription factor 5 (Tbx5 ; avg log fold change 0.54, p= 2.1E-110) in the AVN; and ETS variant transcription factor 1 ( Etv1; avg log fold change 0.89, p= 6.7E-105) in the PFs, among others. Notably, an additional 16 novel DNA-binding transcription factors and 40 transcription co-regulators were enriched within specific CCS components at statistical significance. To further evaluate the transcriptional landscape of individual subcomponents within the SAN (compact versus transitional cells), we also performed scRNAseq on FACS-isolated SAN cells from embryonic day 16.5 Hcn4 -GFP transgenic mice revealing differential expression of a host of TFs, both established and novel. Finally, our murine findings were then cross-referenced to over 15 human GWAS studies related to cardiac rhythm abnormalities, revealing an association of a subset of these murine CCS-enriched transcription factors and cardiac arrhythmias in humans. In conclusion, this study provides a comprehensive gene expression atlas of transcription factors and transcriptional co-regulators within individual CCS components at single-cell resolution, laying the groundwork for future studies into the transcriptional regulation of cardiac conduction development and disease.

A chemotherapy response prediction model derived from tumor-promoting B and Tregs and proinflammatory macrophages in HGSOC.

Most patients with high-grade serous ovarian cancer (HGSOC) experienced disease recurrence with cumulative chemoresistance, leading to treatment failure. However, few biomarkers are currently available in clinical practice that can accurately predict chemotherapy response. The tumor immune microenvironment is critical for cancer development, and its transcriptomic profile may be associated with treatment response and differential outcomes. The aim of this study was to develop a new predictive signature for chemotherapy in patients with HGSOC. Two HGSOC single-cell RNA sequencing datasets from patients receiving chemotherapy were reinvestigated. The subtypes of endoplasmic reticulum stress-related XBP1+ B cells, invasive metastasis-related ACTB+ Tregs, and proinflammatory-related macrophage subtypes with good predictive power and associated with chemotherapy response were identified. These results were verified in an independent HGSOC bulk RNA-seq dataset for chemotherapy. Further validation in clinical cohorts used quantitative real-time PCR (qRT-PCR). By combining cluster-specific genes for the aforementioned cell subtypes, we constructed a chemotherapy response prediction model containing 43 signature genes that achieved an area under the receiver operator curve (AUC) of 0.97 (p = 2.1e-07) for the GSE156699 cohort (88 samples). A huge improvement was achieved compared to existing prediction models with a maximum AUC of 0.74. In addition, its predictive capability was validated in multiple independent bulk RNA-seq datasets. The qRT-PCR results demonstrate that the expression of the six genes has the highest diagnostic value, consistent with the trend observed in the analysis of public data. The developed chemotherapy response prediction model can be used as a valuable clinical decision tool to guide chemotherapy in HGSOC patients.

A new human embryonic cell type associated with activity of young transposable elements allows definition of the inner cell mass.

There remains much that we do not understand about the earliest stages of human development. On a gross level, there is evidence for apoptosis, but the nature of the affected cell types is unknown. Perhaps most importantly, the inner cell mass (ICM), from which the foetus is derived and hence of interest in reproductive health and regenerative medicine, has proven hard to define. Here, we provide a multi-method analysis of the early human embryo to resolve these issues. Single-cell analysis (on multiple independent datasets), supported by embryo visualisation, uncovers a common previously uncharacterised class of cells lacking commitment markers that segregates after embryonic gene activation (EGA) and shortly after undergo apoptosis. The discovery of this cell type allows us to clearly define their viable ontogenetic sisters, these being the cells of the ICM. While ICM is characterised by the activity of an Old non-transposing endogenous retrovirus (HERVH) that acts to suppress Young transposable elements, the new cell type, by contrast, expresses transpositionally competent Young elements and DNA-damage response genes. As the Young elements are RetroElements and the cells are excluded from the developmental process, we dub these REject cells. With these and ICM being characterised by differential mobile element activities, the human embryo may be a "selection arena" in which one group of cells selectively die, while other less damaged cells persist.

Comprehensive analyses of brain cell communications based on multiple scRNA-seq and snRNA-seq datasets for revealing novel mechanism in neurodegenerative diseases.

Complex cellular communications between glial cells and neurons are critical for brain normal function and disorders, and single-cell level RNA-sequencing datasets display more advantages for analyzing cell communications. Therefore, it is necessary to systematically explore brain cell communications when considering factors such as sex and brain region. We extracted a total of 1,039,459 cells derived from 28 brain single-cell RNA-sequencing (scRNA-seq) or single-nucleus RNA-sequencing (snRNA-seq) datasets from the GEO database, including 12 human and 16 mouse datasets. These datasets were further divided into 71 new sub-datasets when considering disease, sex, and region conditions. In the meanwhile, we integrated four methods to evaluate ligand-receptor interaction score among six major brain cell types (microglia, neuron, astrocyte, oligodendrocyte, OPC, and endothelial cell). For Alzheimer's disease (AD), disease-specific ligand-receptor pairs when compared with normal sub-datasets, such as SEMA4A-NRP1, were identified. Furthermore, we explored the sex- and region-specific cell communications and identified that WNT5A-ROR1 among microglia cells displayed close communications in male, and SPP1-ITGAV displayed close communications in the meninges region from microglia to neurons. Furthermore, based on the AD-specific cell communications, we constructed a model for AD early prediction and confirmed the predictive performance using multiple independent datasets. Finally, we developed an online platform for researchers to explore brain condition-specific cell communications. This research provided a comprehensive study to explore brain cell communications, which could reveal novel biological mechanisms involved in normal brain function and neurodegenerative diseases such as AD.

Inferring evolutionary trajectories from cross-sectional transcriptomic data to mirror lung adenocarcinoma progression.

Lung adenocarcinoma (LUAD) is a deadly tumor with dynamic evolutionary process. Although much endeavors have been made in identifying the temporal patterns of cancer progression, it remains challenging to infer and interpret the molecular alterations associated with cancer development and progression. To this end, we developed a computational approach to infer the progression trajectory based on cross-sectional transcriptomic data. Analysis of the LUAD data using our approach revealed a linear trajectory with three different branches for malignant progression, and the results showed consistency in three independent cohorts. We used the progression model to elucidate the potential molecular events in LUAD progression. Further analysis showed that overexpression of BUB1B, BUB1 and BUB3 promoted tumor cell proliferation and metastases by disturbing the spindle assembly checkpoint (SAC) in the mitosis. Aberrant mitotic spindle checkpoint signaling appeared to be one of the key factors promoting LUAD progression. We found the inferred cancer trajectory allows to identify LUAD susceptibility genetic variations using genome-wide association analysis. This result shows the opportunity for combining analysis of candidate genetic factors with disease progression. Furthermore, the trajectory showed clear evident mutation accumulation and clonal expansion along with the LUAD progression. Understanding how tumors evolve and identifying mutated genes will help guide cancer management. We investigated the clonal architectures and identified distinct clones and subclones in different LUAD branches. Validation of the model in multiple independent data sets and correlation analysis with clinical results demonstrate that our method is effective and unbiased.

Characterization of molecular subtypes based on chromatin regulators and identification of the role of NPAS2 in lung adenocarcinoma

BackgroundChromatin regulators (CRs) are critical epigenetic modifiers and have been reported to play critical roles during the progression of various tumors, but their role in lung adenocarcinoma (LUAD) has not been comprehensively studied.MethodsDifferential expression and univariate Cox regression analyses were conducted to identify the prognostic CRs. Consensus clustering was applied to classify the subtypes of LUAD based on prognostic CRs. LASSO-multivariate Cox regression method was used for construction of a prognostic signature and development of chromatin regulator-related gene index (CRGI). The capacity of CRGI to distinguish survival was evaluated via Kaplan–Meier method in multiple datasets. Relationship between CRGI and tumor microenvironment (TME) was evaluated. Additionally, clinical variables and CRGI were incorporated to create a nomogram. The role of the prognostic gene NPAS2 in LUAD was elucidated via clinical samples validation and a series of in vitro and in vivo experiments.ResultsTwo subtypes of LUAD were classified based on 46 prognostic CRs via consensus clustering which had significantly different survival and TME. A prognostic signature consisting of six CRs (MOCS, PBK, CBX3, A1CF, NPAS2, and CTCFL) was developed and proved to be an effective survival predictor in multiple independent datasets. The prognostic signature was also demonstrated to be an indicator of TME and sensitivity to immunotherapy and chemotherapy. The nomogram was suggested to be a simple tool that can predict survival accurately. Clinical samples show that NPAS2 is highly expressed in LUAD tissues, and in vitro and in vivo experiments demonstrated that inhibition of NPAS2 impeded malignant progression of LUAD cells.ConclusionsOur study comprehensively unveiled the functions of CRs in LUAD, developed a classifier to predict survival and response to treatments, and suggested that NPAS2 promoted LUAD progression for the first time.

Construction and evaluation of Alzheimer's disease diagnostic prediction model based on genes involved in mitophagy.

Alzheimer's disease (AD) is a common neurodegenerative disease. The concealment of the disease is the difficulty of its prevention and treatment. Previous studies have shown that mitophagy is crucial to the development of AD. However, there is a lack of research on the identification and clinical significance of mitophagy-related genes in AD. Therefore, the purpose of this study was to identify the mitophagy-related genes with the diagnostic potential for AD and establish a diagnostic model for AD. Firstly, we download the AD gene expression profile from Gene Expression Omnibus (GEO). Limma, PPI, functional enrichment analysis and WGCNA were used to screen the differential expression of mitophagy-related AD gene. Then, machine learning methods (random forest, univariate analysis, support vector machine, LASSO regression and support vector machine classification) were used to identify diagnostic markers. Finally, the diagnostic model was established and evaluated by ROC, multiple regression analysis, nomogram, calibration curve and other methods. Moreover, multiple independent datasets, AD cell models and AD clinical samples were used to verify the expression level of characteristic genes in the diagnostic model. In total, 72 differentially expressed mitophagy-related related genes were identified, which were mainly involved in biological functions such as autophagy, apoptosis and neurological diseases. Four mitophagy-related genes (OPTN, PTGS2, TOMM20, and VDAC1) were identified as biomarkers. A diagnostic prediction model was constructed, and the reliability of the model was verified by receiver operating characteristic (ROC) curve analysis of GSE122063 and GSE63061. Then, we combine four mitophagy-related genes with age to establish a nomogram model. The ROC, C index and calibration curve show that the model has good prediction performance. Finally, multiple independent datasets, AD cell model samples and clinical peripheral blood samples confirmed that the expression levels of four mitophagy-related genes were consistent with the results of bioinformatics analysis. The analysis results and diagnostic model of this study are helpful for the follow-up clinical work and mechanism research of AD.

The integrated single-cell analysis developed a lactate metabolism-driven signature to improve outcomes and immunotherapy in lung adenocarcinoma.

It has been suggested that lactate metabolism (LM) is crucial for the development of cancer. Using integrated single-cell RNA sequencing (scRNA-seq) analysis, we built predictive models based on LM-related genes (LMRGs) to propose novel targets for the treatment of LUAD patients. The most significant genes for LM were identified through the use of the AUCell algorithm and correlation analysis in conjunction with scRNA-seq analysis. To build risk models with superior predictive performance, cox- and lasso-regression were utilized, and these models were validated on multiple external independent datasets. We then explored the differences in the tumor microenvironment (TME), immunotherapy, mutation landscape, and enriched pathways between different risk groups. Finally, cell experiments were conducted to verify the impact of AHSA1 in LUAD. A total of 590 genes that regulate LM were identified for subsequent analysis. Using cox- and lasso-regression, we constructed a 5-gene signature that can predict the prognosis of patients with LUAD. Notably, we observed differences in TME, immune cell infiltration levels, immune checkpoint levels, and mutation landscapes between different risk groups, which could have important implications for the clinical treatment of LUAD patients. Based on LMRGs, we constructed a prognostic model that can predict the efficacy of immunotherapy and provide a new direction for treating LUAD.

Phylogenomics reveals within species diversification but incongruence with color phenotypes in widespread orchid bees (Hymenoptera: Apidae: Euglossini)

AbstractColoration is an important phenotypic trait for taxonomic studies and has been widely used for identifying insect species and populations. However, coloration can be a poor diagnostic character for insect species that exhibit high polymorphism in this trait, which can lead to over-splitting of taxonomic units. In orchid bees, color variation has been interpreted by different taxonomists as either polymorphism associated with Müllerian mimicry complexes or diagnostic traits for species identification. Despite this uncertainty, integrative approaches that incorporate multiple independent datasets to test the validity of hair coloration as a character that identifies independent evolutionary units have not been used. Here, we use phylogenomic data from Ultraconserved Elements (UCEs) to explore whether color phenotypes in the widespread orchid bee species complexes Eulaema meriana and Eulaema bombiformis (Hymenoptera: Apidae: Euglossini) correspond to independent lineages or polymorphic trait variation within species. We find that lineages within both species are structured according to geography and that color morphs are generally unassociated with evolutionarily independent groups except for populations located in the Atlantic Forest of Brazil. We conclude that there is compelling evidence that E. atleticana and E. niveofasciata are subspecies of E. meriana and E. bombiformis, respectively, and not different species as previously suggested. Therefore, we recognize Eulaema meriana atleticanacomb. n. and Eulaema bombiformis niveofasciatacomb. n. and discuss their morphological characteristics. We make recommendations on the use of color traits for orchid bee taxonomy and discuss the significance of subspecies as evolutionary units relevant for conservation efforts.

False discovery rate-controlled multiple testing for union null hypotheses: a knockoff-based approach.

False discovery rate (FDR) controlling procedures provide important statistical guarantees for replicability in signal identification based on multiple hypotheses testing. In many fields of study, FDR controling procedures are used in high-dimensional (HD) analyses to discover features that are truly associated with the outcome. In some recent applications, data on the same set of candidate features are independently collected in multiple different studies. For example, gene expression data are collected at different facilities and with different cohorts, to identify the genetic biomarkers of multiple types of cancers. These studies provide us with opportunities to identify signals by considering information from different sources (with potential heterogeneity) jointly. This paper is about how to provide FDR control guarantees for the tests of union null hypotheses of conditional independence. We present a knockoff-based variable selection method (Simultaneous knockoffs) to identify mutual signals from multiple independent datasets, providing exact FDR control guarantees under finite sample settings. This method can work with very general model settings and test statistics. We demonstrate the performance of this method with extensive numerical studies and two real-dataexamples.

ATP8B1: A prognostic prostate cancer biomarker identified via genetic analysis.

Controlling the asymmetric distribution of phospholipids across biological membranes plays a pivotal role in the life cycle of cells; one of the most important contributors that maintain this lipid asymmetry are phospholipid-transporting adenosine triphosphatases (ATPases). Although sufficient information regarding their association with cancer exists, there is limited evidence linking the genetic variants of phospholipid-transporting ATPase family genes to prostate cancer in humans. In this study, we investigated the association of 222 haplotype-tagging single-nucleotide polymorphisms (SNPs) in eight phospholipid-transporting ATPase genes with cancer-specific survival (CSS) and overall survival (OS) of 630 patients treated with androgen-deprivation therapy (ADT) for prostate cancer. After multivariate Cox regression analysis and multiple testing correction, we found that ATP8B1 rs7239484 was remarkably associated with CSS and OS after ADT. A pooled analysis of multiple independent gene-expression datasets demonstrated that ATP8B1 was under-expressed in tumor tissues and that a higher ATP8B1 expression was associated with a better patient prognosis. Moreover, we established highly invasive sublines using two human prostate cancer cell lines to mimic cancer progression traits in vitro. The expression of ATP8B1 was consistently downregulated in both highly invasive sublines. Our study indicates that rs7239484 is a prognostic factor for patients treated with ADT and that ATP8B1 can potentially attenuate prostate cancer progression.

Establishing conversion of the 16-item Informant Questionnaire on Cognitive Decline in the Elderly scores into interval-level data across multiple samples using Rasch methodology.

The 16-item Informant Questionnaire on Cognitive Decline in the Elderly (IQCODE-16) is a well-validated and widely-used measure of cognitive changes (CCs) among older adults. This study aimed to use Rasch methodology to establish psychometric properties of the IQCODE-16 and validate the existing ordinal-to-interval transformation algorithms across multiple large samples. A Partial Credit Rasch model was employed to examine psychometric properties of the IQCODE-16 using data (n=918) from two longitudinal studies of participants aged 57-99 years: the Older Australian Twins Study (n=450) and the Canberra Longitudinal Study (n=468), and reusing the Sydney Memory and Ageing Study (MAS) sample (n=400). Initial analyses indicated good reliability for the IQCODE-16 (Person Separation Index range: 0.82-0.90). However, local dependency was identified between items, with several items showing misfit to the model. Replicating the existing Rasch solution could not reproduce the best Rasch model fit for all samples. Combining locally dependent items into three testlets resolved all misfit and local dependency issues and resulted in the best Rasch model fit for all samples with evidence of unidimensionality, strong reliability, and invariance across person factors. Accordingly, new ordinal-to-interval transformation algorithms were produced to convert the IQCODE-16 ordinal scores into interval data to improve the accuracy of its scores. The findings of this study support the reliability and validity of the IQCODE-16 in measuring CCs among older adults. New ordinal-to-interval conversion tables generated using samples from multiple independent datasets are more generalizable and can be used to enhance the precision of the IQCODE-16 without changing its original format. An easy-to-use converter has been made available for clinical and research use.

Estrogen-related genes influence immune cell infiltration and immunotherapy response in Hepatocellular Carcinoma.

Immunotherapy has been the first-line treatment option in advanced Hepatocellular Carcinoma(HCC); but now, there are no established molecular markers that can predict immunotherapy response. Estrogen has a crucial role in the development of a variety of liver illnesses, including liver fibrosis, Nonalcoholic fatty liver disease (NAFLD), and HCC. Nonetheless, the significance of estrogen-related genes in HCC immunotherapy and the underlying molecular mechanisms are not yet fully understood. In this study, we constructed a novel estrogen-related gene prognostic signature (ERGPS) by analyzing bulk RNA sequencing data from 365 HCC patients. Based on the median risk score, we divided 365 HCC patients into low- and high-risk groups. Tumor mutation burden (TMB), Microsatellite instability (MSI), T cell receptor (TCR) richness, B cell receptor (BCR) richness, single-nucleotide variants (SNV) Neoantigens, Cancer Testicular Antigens (CTA) scores, and Tumour Immune Dysfunction and Exclusion (TIDE) scores were used to evaluate the magnitude of immunotherapy response. Multiple external datasets validate the validity and robustness of the prognostic signature. Real-time quantitative polymerase chain reaction (qRT-PCR) was used to validate estrogen-related gene overexpression in HCC tissue samples. ERGPS is an independent risk factor affecting the prognosis of HCC patients and is superior to other clinical variables in predicting patient survival and immunotherapy response. Multiple independent external datasets confirmed the superior predictive efficacy of the prognostic signature. The prognostic signature was positively correlated with TMB score, MSI score, TCR richness, BCR richness, SNV Neoantigens score, CTA score, expression levels of immune checkpoint-related genes, and TIDE score. Patients with HCC in the high-risk group identified by the prognostic signature were likely to be more responsive to immunotherapy and more suitable for immunotherapy. qRT-PCR confirmed that estrogen-related genes of the construct signature were highly expressed in HCC tumor tissues. Estrogen-related genes are overexpressed in HCC tissues. Our novel prognostic signature can accurately predict not only the prognosis but also the immunotherapy response of HCC patients. In the future, prognostic signatures will be a useful tool for clinicians to screen patients with HCC who are suitable for immunotherapy.

MIMESIS: minimal DNA-methylation signatures to quantify and classify tumor signals in tissue and cell-free DNA samples.

DNA-methylation alterations are common in cancer and display unique characteristics that make them ideal markers for tumor quantification and classification. Here we present MIMESIS, a computational framework exploiting minimal DNA-methylation signatures composed by a few dozen informative DNA-methylation sites to quantify and classify tumor signals in tissue and cell-free DNA samples. Extensive analyses of multiple independent and heterogenous datasets including >7200 samples demonstrate the capability of MIMESIS to provide precise estimations of tumor content and to enable accurate classification of tumor type and molecular subtype. To assess our framework for clinical applications, we designed a MIMESIS-informed assay incorporating the minimal signatures for breast cancer. Using both artificial samples and clinical serial cell-free DNA samples from patients with metastatic breast cancer, we show that our approach provides accurate estimations of tumor content, sensitive detection of tumor signal and the ability to capture clinically relevant molecular subtype in patients' circulation. This study provides evidence that our extremely parsimonious approach can be used to develop cost-effective and highly scalable DNA-methylation assays that could support and facilitate the implementation of precision oncology in clinical practice.

PIQLE: protein-protein interface quality estimation by deep graph learning of multimeric interaction geometries.

Accurate modeling of protein-protein interaction interface is essential for high-quality protein complex structure prediction. Existing approaches for estimating the quality of a predicted protein complex structural model utilize only the physicochemical properties or energetic contributions of the interacting atoms, ignoring evolutionarily information or inter-atomic multimeric geometries, including interaction distance and orientations. Here, we present PIQLE, a deep graph learning method for protein-protein interface quality estimation. PIQLE leverages multimeric interaction geometries and evolutionarily information along with sequence- and structure-derived features to estimate the quality of individual interactions between the interfacial residues using a multi-head graph attention network and then probabilistically combines the estimated quality for scoring the overall interface. Experimental results show that PIQLE consistently outperforms existing state-of-the-art methods including DProQA, TRScore, GNN-DOVE and DOVE on multiple independent test datasets across a wide range of evaluation metrics. Our ablation study and comparison with the self-assessment module of AlphaFold-Multimer repurposed for protein complex scoring reveal that the performance gains are connected to the effectiveness of the multi-head graph attention network in leveraging multimeric interaction geometries and evolutionary information along with other sequence- and structure-derived features adopted in PIQLE. An open-source software implementation of PIQLE is freely available at https://github.com/Bhattacharya-Lab/PIQLE. Supplementary data are available at Bioinformatics Advances online.

Comprehensive analysis of hypoxia-related genes for prognosis, immune features, and drugs treatment strategy in gastric cancer using bulk and single-cell RNA-sequencing

Hypoxia is one of the malignant characteristics of solid tumors and is related to the multiple malignant characteristics of the tumor. No study has not yet reported a systematical analysis of the characteristics of hypoxia from single-cell resolution in gastric cancer. In our research, we investigated the hypoxia features of various types of cells in single-cell resolution, identified hypoxia-related genes by the weighted gene co-expression network analysis method. Through the hypoxia-related genes from single-cell levels, we screened out 13 genes and established a prognostic model. This model performs well in the training dataset and multiple independent verification data sets. We thought that tumor hypoxia might affect the DNA methylation of cells and promote the transcription of genes associated with malignant features, thereby promoting tumor progression. We found that the more tumor associated genes in the high-risk group showed hypomethylation and high hypoxia-risk score group have more tumor-related genes, more immunosuppressive immune cells and more enrichment of cancer -related pathways. The lower risk group is more sensitive to three chemotherapy drugs for gastric cancer. Our study illustrates the crucial role of hypoxia in gastric cancer. Hypoxia-related gene prognostic model has been established and has good performance. Hypoxia-related risk score can also be used to guide a patient’s drug treatment strategy.

Identifying risk genes and therapeutics for Alzheimer’s disease through integration multi‐omics and drug‐perturbation signatures

AbstractBackgroundAlzheimer’s disease (AD) is a severe and irreversible neurodegenerative disease with no effective treatments currently availableMethodTo better understand AD mechanisms and identify therapeutics, we collected AD GWAS loci from the largest meta‐analysis and incorporated multiple lines of genomic evidence, including human brain chromatin interaction map in microglia, brain cell‐type specific expression and gene‐gene interaction network, to infer high‐confidence risk genes (HRGs) at each GWAS locus. We then used the HRGs and network propagation to construct a genetic perturbation signature of AD, and queried a database of drug‐induced transcriptomic signatures (LINCS 1000) to evaluate compounds that are able to reverse the genetic perturbation signature.ResultThis resulted in 72 HRGs, which were further validated in multiple independent datasets of human microglia expression quantitative trait loci (eQTLs) and splicing quantitative trait loci (sQTLs), protein quantitative trait loci (pQTLs) in human brain, cerebrospinal fluid (CSF) and plasma, and genes of protein targets in brain and CSF associated with asymptomatic AD. Pathway analysis implicates the involvement of endocytosis and amyloid pathways, and highlights cellular compartments such as gamma‐secretase complex, NF‐kappaB complex, neurofibrillary tangle and endosome. Query of the genetic perturbation signature of AD against LINCS 1000 revealed several drugs that can reverse the genetic perturbation signature.ConclusionIn conclusion, by incorporating multiple lines of evidence and resources, our study offers new insights into the pathological processes underlying AD and potential therapeutic drugs.

Deep learning model to identify angle dysgenesis on ASOCT images

AbstractAnterior segment spectral domain–optical coherence tomography (ASOCT) is increasingly used in glaucoma patients, primarily to investigate the anterior chamber angle and visualize the trabecular meshwork (TM) and Schlemm's Canal (SC) in‐vivo. The anatomical variability of the drainage angle, TM and SC morphology makes it difficult to distinguish normal from abnormal. Therefore, the interpretation of ASOCT images requires expertise and a deep understanding of the complexity involved in the developmental anomalies of the outflow pathways. While the ASOCT scans can successfully capture the gross changes, they may fail to identify angle dysgenesis in cases where altered extracellular matrix anomalies are subtle enough to get unnoticed by the human eyes of an expert. Deep learning (DL) has demonstrated convincing results and potential in classifying ophthalmic diseases. The integration of high‐resolution ASOCT images trained DL algorithms can further enhance the translational potential of DL‐based classification for angle dysgenesis. Apart from the gross changes in the iridocorneal angle region, the perturbations in TM layers due to associated gene mutations are also captured on ASOCT scans and may be identified by DL algorithms. We have developed DL models, which can identify obscure features of angle dysgenesis in ASOCT scans (ADoA) and validated its performance in eyes with primary open angle glaucoma. Eight hundred ASOCT B‐scans were included, and each image was cropped to extract the iridio‐corneal angle (ICA) area and a TM area to form two datasets. Finally, the total number of images after augmentations in each dataset increased to 8160, divided into training and testing sets for model building. To avoid overfitting, incorporate variability and increase the size of the datasets, 23 different types of augmentation techniques were performed on each image using the CloDSA tool, a freely available package.Three deep learning models were developed for a consensus‐based outcome to identify ADoA objectively. The models performed well on the internal as well as external validation datasets. The model performance on the multiple independent datasets highlights the translational relevance of image‐based diagnostic alternatives in clinical settings, as it has the potential to be deployed for screening patients and their family members for detection of angle dysgenesis in vivo.

Highly expressed FAM189B predicts poor prognosis in hepatocellular carcinoma.

Hepatocellular carcinoma (HCC) is one of the most malignant tumors with persistently high morbidity and mortality. However, the expression, prognostic and clinical significance of FAM189 family genes in HCC remain largely unknown. In this study, the expression levels of FAM189 family genes in HCC were analyzed through TCGA-LIHC and ICGC-LIRI-JP cohorts, and further validated in multiple independent GEO datasets. It was found that the expression of FAM189B was significantly upregulated in HCC tumor tissues, while the expression of FAM189A1 and FAM189A2 was not significantly changed between tumor and adjacent tissues. Further analysis revealed that upregulated copy number variation contributed to increased expression of FAM189B in HCC. Survival analysis showed that highly expressed FAM189B was significantly correlated with unfavorable prognosis, including overall survival, disease-specific survival, and progression-free interval. Univariate and multivariate Cox regression analysis showed that FAM189B was a potential novel prognosis factor for HCC patients. In addition, the association between FAM189B expression and clinical and molecular characteristics was analyzed. High expression of FAM189B was associated with high AFP level, high predicted risk metastasis signature, and TP53 mutation, while there was no significant association between FAM189B expression and cancer stage or tumor grade of HCC. Gene set enrichment analysis revealed that highly expressed FAM189B was closely related with signal pathways and biological processes associated with cell proliferation and cell cycle in HCC. In conclusion, this study suggested that FAM189B was highly expressed in HCC and highly expressed FAM189B may serve as an effective prognostic indicator and a potential therapeutic target for HCC patients.

A novel oxidative stress- and ferroptosis-related gene prognostic signature for distinguishing cold and hot tumors in colorectal cancer.

Oxidative stress and ferroptosis exhibit crosstalk in many types of human diseases, including malignant tumors. We aimed to develop an oxidative stress- and ferroptosis-related gene (OFRG) prognostic signature to predict the prognosis and therapeutic response in patients with colorectal cancer (CRC). Thirty-four insertion genes between oxidative stress-related genes and ferroptosis-related genes were identified as OFRGs. We then performed bioinformatics analysis of the expression profiles of 34 OFRGs and clinical information of patients obtained from multiple datasets. Patients with CRC were divided into three OFRG clusters, and differentially expressed genes (DEGs) between clusters were identified. OFRG clusters correlated with patient survival and immune cell infiltration. Prognosis-related DEGs in three clusters were used to calculate the risk score, and a prognostic signature was constructed according to the risk score. In this study, patients in the low-risk group had better prognosis, higher immune cell infiltration levels, and better responses to fluorouracil-based chemotherapy and immune checkpoint blockade therapy than high-risk patients; these results were successfully validated with multiple independent datasets. Thus, low-risk CRC could be defined as hot tumors and high-risk CRC could be defined as cold tumors. To further identify potential biomarkers for CRC, the expression levels of five signature genes in CRC and adjacent normal tissues were further verified via an in vitro experiment. In conclusion, we identified 34 OFRGs and constructed an OFRG-related prognostic signature, which showed excellent performance in predicting survival and therapeutic responses for patients with CRC. This could help to distinguish cold and hot tumors in CRC, and the results might be helpful for precise treatment protocols in clinical practice.