CROST: a comprehensive repository of spatial transcriptomics.

The development of spatial transcriptome sequencing technology has revolutionized our comprehension of complex tissues and propelled life and health sciences into an era of spatial omics. However, the current availability of databases for accessing and analyzing spatial transcriptomic data is limited. In response, we have established CROST (https://ngdc.cncb.ac.cn/crost), a comprehensive repository of spatial transcriptomics. CROST encompasses high-quality samples and houses 182 spatial transcriptomic datasets from diverse species, organs, and diseases, comprising 1033 sub-datasets and 48 043 tumor-related spatially variable genes (SVGs). Additionally, it encompasses a standardized spatial transcriptome data processing pipeline, integrates single-cell RNA sequencing deconvolution spatial transcriptomics data, and evaluates correlation, colocalization, intercellular communication, and biological function annotation analyses. Moreover, CROST integrates the transcriptome, epigenome, and genome to explore tumor-associated SVGs and provides a comprehensive understanding of their roles in cancer progression and prognosis. Furthermore, CROST provides two online tools, single-sample gene set enrichment analysis and SpatialAP, for users to annotate and analyze the uploaded spatial transcriptomics data. The user-friendly interface of CROST facilitates browsing, searching, analyzing, visualizing, and downloading desired information. Collectively, CROST offers fresh and comprehensive insights into tissue structure and a foundation for understanding multiple biological mechanisms in diseases, particularly in tumor tissues.

Transcriptomic and genomic characteristics of intrahepatic metastases of primary liver cancer.

BACKGROUND: Patients with primary multifocal hepatocellular carcinoma (HCC) have a poor prognosis and often experience a high rate of treatment failure. Multifocal HCC is mainly caused by intrahepatic metastasis (IM), and though portal vein tumor thrombosis (PVTT) is considered a hallmark of IM, the molecular mechanism by which primary HCC cells invade the portal veins remains unclear. Therefore, it is necessary to recognize the early signs of metastasis of HCC to arrange better treatment for patients. RESULTS: To determine the differential molecular features between primary HCC with and without phenotype of metastasis, we used the CIBERSORTx software to deconvolute cell types from bulk RNA-Seq based on a single-cell transcriptomic dataset. According to the relative abundance of tumorigenic and metastatic hepatoma cells, VEGFA+ macrophages, effector memory T cells, and natural killer cells, HCC samples were divided into five groups: Pro-T, Mix, Pro-Meta, NKC, and MemT, and the transcriptomic and genomic features of the first three groups were analyzed. We found that the Pro-T group appeared to retain native hepatic metabolic activity, whereas the Pro-Meta group underwent dedifferentiation. Genes highly expressed in the group Pro-Meta often signify a worse outcome. CONCLUSIONS: The HCC cohort can be well-typed and prognosis predicted according to tumor microenvironment components. Primary hepatocellular carcinoma may have obtained corresponding molecular features before metastasis occurred.

DNA methylation-based age prediction with bloodstains using pyrosequencing and random forest regression.

The use of DNA methylation to predict chronological age has shown promising potential for obtaining additional information in forensic investigations. To date, several studies have reported age prediction models based on DNA methylation in body fluids with high DNA content. However, it is often difficult to apply these existing methods in practice due to the low amount of DNA present in stains of body fluids that are part of a trace material. In this study, we present a sensitive and rapid test for age prediction with bloodstains based on pyrosequencing and random forest regression. This assay requires only 0.1 ng of genomic DNA and the entire procedure can be completed within 10 h, making it practical for forensic investigations that require a short turnaround time. We examined the methylation levels of 46 CpG sites from six genes using bloodstain samples from 128 males and 113 females aged 10-79 years. A random forest regression model was then used to construct an age prediction model for males and females separately. The final age prediction models were developed with seven CpG sites (three for males and four for females) based on the performance of the random forest regression. The mean absolute deviation was less than 3 years for each model. Our results demonstrate that DNA methylation-based age prediction using pyrosequencing and random forest regression has potential applications in forensics to accurately predict the biological age of a bloodstain donor.

Overview of multi-omics research in gout.

Gout is a self-limiting inflammation disease triggered by deposition of monosodium urate with a variety of comorbidities. With the improvement of living standards, the global incidence of gout is increasing year by year, which seriously affects people's health. As an effective tool to study diseases, omics technology has been widely used to discover potential biomarkers and risk factors of gout. The identified variation sites or different-expressed products provide different dimensions of insights for the study of the pathogenesis and disease progression of gout. In this review, the application and research results of multi-omics technology in gout were analyzed and summarized through PubMed literature retrieval. Meanwhile, the recent research progress of multi-omics technology in the field of gout was reviewed to understand the specific changes of gout patients at different molecular levels, and to provide ideas and directions for further research on gout in the future.

Platform-independent approach for cancer detection from gene expression profiles of peripheral blood cells.

Peripheral blood gene expression intensity-based methods for distinguishing healthy individuals from cancer patients are limited by sensitivity to batch effects and data normalization and variability between expression profiling assays. To improve the robustness and precision of blood gene expression-based tumour detection, it is necessary to perform molecular diagnostic tests using a more stable approach. Taking breast cancer as an example, we propose a machine learning-based framework that distinguishes breast cancer patients from healthy subjects by pairwise rank transformation of gene expression intensity in each sample. We showed the diagnostic potential of the method by performing RNA-seq for 37 peripheral blood samples from breast cancer patients and by collecting RNA-seq data from healthy donors in Genotype-Tissue Expression project and microarray mRNA expression datasets in Gene Expression Omnibus. The framework was insensitive to experimental batch effects and data normalization, and it can be simultaneously applied to new sample prediction.

Omics big data and medical artificial intelligence.

With the rapid development of high-throughput sequencing technology and computer science, the amount of large omics data has increased exponentially, the advantages of multi-omics analysis have gradually emerged, and the application of artificial intelligence has become more and more extensive. In this review, we introduce the application progress of multi-omics data analysis and artificial intelligence in the medical field in recent years, and also show the cases and advantages of their combined application. Finally, we briefly explain the current challenges of multi-omics analysis and artificial intelligence in order to provide new research ideas for the medical industry and to promote the development and application of precision medicine.

The accessible chromatin landscape of the human genome.

DNase I hypersensitive sites (DHSs) are markers of regulatory DNA and have underpinned the discovery of all classes of cis-regulatory elements including enhancers, promoters, insulators, silencers and locus control regions. Here we present the first extensive map of human DHSs identified through genome-wide profiling in 125 diverse cell and tissue types. We identify ∼2.9 million DHSs that encompass virtually all known experimentally validated cis-regulatory sequences and expose a vast trove of novel elements, most with highly cell-selective regulation. Annotating these elements using ENCODE data reveals novel relationships between chromatin accessibility, transcription, DNA methylation and regulatory factor occupancy patterns. We connect ∼580,000 distal DHSs with their target promoters, revealing systematic pairing of different classes of distal DHSs and specific promoter types. Patterning of chromatin accessibility at many regulatory regions is organized with dozens to hundreds of co-activated elements, and the transcellular DNase I sensitivity pattern at a given region can predict cell-type-specific functional behaviours. The DHS landscape shows signatures of recent functional evolutionary constraint. However, the DHS compartment in pluripotent and immortalized cells exhibits higher mutation rates than that in highly differentiated cells, exposing an unexpected link between chromatin accessibility, proliferative potential and patterns of human variation.

MicroRNA 200a inhibits erythroid differentiation by targeting PDCD4 and THRB.

Previous studies on erythropoiesis revealed that microRNAs (miRNAs) play a critical role in erythroid differentiation. Given the abundance of identified miRNAs and the limited understanding of erythroid miRNAs, additional examination is required. Here, two sets of erythroid differentiation miRNome data were analysed to screen for novel erythroid-inhibiting miRNAs. MIR200A was selected based on its pattern of downregulated expression in the miRNome datasets during induction of erythroid differentiation. Overexpression of MIR200A in K562 and TF-1 cells confirmed its inhibitory role in erythroid differentiation. Further in vivo study indicated that overexpression of mir200a inhibited primitive erythropoiesis of zebrafish. Transcriptome analyses after MIR200A overexpression in TF-1 cells indicated a significant role in regulating erythroid function and revealed potential regulation networks. Additionally, bioinformatics and experimental analyses confirmed that PDCD4 (programmed cell death 4) and THRB (thyroid hormone receptor, beta) are both targets of MIR200A-3p. Gain- and loss-of-function studies of PDCD4 and THRB revealed that the two targets were capable of promoting erythroid gene expression. Overall, our results revealed that microRNA 200a inhibits erythroid differentiation by targeting PDCD4 and THRB.

Knockdown of transcription factor forkhead box O3 (FOXO3) suppresses erythroid differentiation in human cells and zebrafish.

Our previous study on the dynamic transcriptomes activated during human erythropoiesis suggested that transcription factor forkhead box O3 (FOXO3) possibly plays a role in erythroid differentiation. Functional studies in human cell line TF-1 indicated that FOXO3 knockdown repressed erythropoietin (EPO)-induced erythroid differentiation by activating promoter region of B-cell translocation gene 1 (BTG1), thereby regulating its expression. In zebrafish, injection of foxo3b-specific morpholinos (foxo3b MO) resulted in reduced globin (hbae1 and hbbe2) and gata1 gene expression. Transcriptome analyses of erythroid lineage cells isolated from the control and foxo3b morphants revealed the dynamic regulation of foxo3b. Further study suggested that BTG1 is partially responsible for FOXO3 regulation in erythroid differentiation of TF-1 cells but is inconsequential in zebrafish. Taken together, we found that FOXO3 plays an important role in erythroid differentiation in both human TF-1 cells and zebrafish, but the mechanism underlying this regulation still remains unclear.

Widespread plasticity in CTCF occupancy linked to DNA methylation.

CTCF is a ubiquitously expressed regulator of fundamental genomic processes including transcription, intra- and interchromosomal interactions, and chromatin structure. Because of its critical role in genome function, CTCF binding patterns have long been assumed to be largely invariant across different cellular environments. Here we analyze genome-wide occupancy patterns of CTCF by ChIP-seq in 19 diverse human cell types, including normal primary cells and immortal lines. We observed highly reproducible yet surprisingly plastic genomic binding landscapes, indicative of strong cell-selective regulation of CTCF occupancy. Comparison with massively parallel bisulfite sequencing data indicates that 41% of variable CTCF binding is linked to differential DNA methylation, concentrated at two critical positions within the CTCF recognition sequence. Unexpectedly, CTCF binding patterns were markedly different in normal versus immortal cells, with the latter showing widespread disruption of CTCF binding associated with increased methylation. Strikingly, this disruption is accompanied by up-regulation of CTCF expression, with the result that both normal and immortal cells maintain the same average number of CTCF occupancy sites genome-wide. These results reveal a tight linkage between DNA methylation and the global occupancy patterns of a major sequence-specific regulatory factor.

Molecular biomarkers screened by next-generation RNA sequencing for non-sentinel lymph node status prediction in breast cancer patients with metastatic sentinel lymph nodes.

BACKGROUND: Non-sentinel lymph node (NSLN) status prediction with molecular biomarkers may make some sentinel lymph node (SLN) positive breast cancer patients avoid the axillary lymph node dissection, but the available markers remain limited. METHODS: SLN positive patients with and without NSLN invasion were selected, and genes differentially expressed or fused in SLN metastasis were screened by next-generation RNA sequencing. RESULTS: Six candidates (all ER/PR+, HER2-, Ki-67 <20%) with metastatic SLNs selected from 305 patients were equally categorized as NSLN negative and positive. We identified 103 specifically expressed genes in the NSLN negative group and 47 in the NSLN positive group. Among them, FABP1 (negative group) and CYP2A13 (positive group) were the only 2 protein-encoding genes with expression levels in the 8th to 10th deciles. Using a false discovery rate threshold of <0.05, 62 up-regulated genes and 98 down-regulated genes were discovered in the NSLN positive group. Furthermore, 10 gene fusions were identified in this group with the most frequently fused gene being IGLL5. CONCLUSIONS: The biomarkers screened in present study may broaden our understanding of the mechanisms of breast cancer metastasis to the lymph nodes and contribute to the axillary surgery selection for SLN positive patients.

MiR-218 Inhibits Erythroid Differentiation and Alters Iron Metabolism by Targeting ALAS2 in K562 Cells.

microRNAs (miRNAs) are involved in a variety of biological processes. The regulatory function and potential role of miRNAs targeting the mRNA of the 5'-aminolevulinate synthase 2 (ALAS2) in erythropoiesis were investigated in order to identify miRNAs which play a role in erythroid iron metabolism and differentiation. Firstly, the role of ALAS2 in erythroid differentiation and iron metabolism in human erythroid leukemia cells (K562) was confirmed by ALAS2 knockdown. Through a series of screening strategies and experimental validations, it was identified that hsa-miR-218 (miR-218) targets and represses the expression of ALAS2 by binding to the 3'-untranslated region (UTR). Overexpression of miR-218 repressed erythroid differentiation and altered iron metabolism in K562 cells similar to that seen in the ALAS2 knockdown in K562 cells. In addition to iron metabolism and erythroid differentiation, miR-218 was found to be responsible for a reduction in K562 cell growth. Taken together, our results show that miR-218 inhibits erythroid differentiation and alters iron metabolism by targeting ALAS2 in K562 cells.

Transcriptome dynamics during human erythroid differentiation and development.

To explore the mechanisms controlling erythroid differentiation and development, we analyzed the genome-wide transcription dynamics occurring during the differentiation of human embryonic stem cells (HESCs) into the erythroid lineage and development of embryonic to adult erythropoiesis using high throughput sequencing technology. HESCs and erythroid cells at three developmental stages: ESER (embryonic), FLER (fetal), and PBER (adult) were analyzed. Our findings revealed that the number of expressed genes decreased during differentiation, whereas the total expression intensity increased. At each of the three transitions (HESCs-ESERs, ESERs-FLERs, and FLERs-PBERs), many differentially expressed genes were observed, which were involved in maintaining pluripotency, early erythroid specification, rapid cell growth, and cell-cell adhesion and interaction. We also discovered dynamic networks and their central nodes in each transition. Our study provides a fundamental basis for further investigation of erythroid differentiation and development, and has implications in using ESERs for transfusion product in clinical settings.

[The ENCODE project and functional genomics studies].

Upon the completion of the Human Genome Project, scientists have been trying to interpret the underlying genomic code for human biology. Since 2003, National Human Genome Research Institute (NHGRI) has invested nearly $0.3 billion and gathered over 440 scientists from more than 32 institutions in the United States, China, United Kingdom, Japan, Spain and Singapore to initiate the Encyclopedia of DNA Elements (ENCODE) project, aiming to identify and analyze all regulatory elements in the human genome. Taking advantage of the development of next-generation sequencing technologies and continuous improvement of experimental methods, ENCODE had made remarkable achievements: identified methylation and histone modification of DNA sequences and their regulatory effects on gene expression through altering chromatin structures, categorized binding sites of various transcription factors and constructed their regulatory networks, further revised and updated database for pseudogenes and non-coding RNA, and identified SNPs in regulatory sequences associated with diseases. These findings help to comprehensively understand information embedded in gene and genome sequences, the function of regulatory elements as well as the molecular mechanism underlying the transcriptional regulation by noncoding regions, and provide extensive data resource for life sciences, particularly for translational medicine. We re-viewed the contributions of high-throughput sequencing platform development and bioinformatical technology improve-ment to the ENCODE project, the association between epigenetics studies and the ENCODE project, and the major achievement of the ENCODE project. We also provided our prospective on the role of the ENCODE project in promoting the development of basic and clinical medicine.

Comprehensive characterization of erythroid-specific enhancers in the genomic regions of human Krüppel-like factors.

BACKGROUND: Mapping of DNase I hypersensitive sites (DHSs) is a powerful tool to experimentally identify cis-regulatory elements (CREs). Among CREs, enhancers are abundant and predominantly act in driving cell-specific gene expression. Krüppel-like factors (KLFs) are a family of eukaryotic transcription factors. Several KLFs have been demonstrated to play important roles in hematopoiesis. However, transcriptional regulation of KLFs via CREs, particularly enhancers, in erythroid cells has been poorly understood. RESULTS: In this study, 23 erythroid-specific or putative erythroid-specific DHSs were identified by DNase-seq in the genomic regions of 17 human KLFs, and their enhancer activities were evaluated using dual-luciferase reporter (DLR) assay. Of the 23 erythroid-specific DHSs, the enhancer activities of 15 DHSs were comparable to that of the classical enhancer HS2 in driving minimal promoter (minP). Fifteen DHSs, some overlapping those that increased minP activities, acted as enhancers when driving the corresponding KLF promoters (KLF-Ps) in erythroid cells; of these, 10 DHSs were finally characterized as erythroid-specific KLF enhancers. These 10 erythroid-specific KLF enhancers were further confirmed using chromatin immunoprecipitation coupled to sequencing (ChIP-seq) data-based bioinformatic and biochemical analyses. CONCLUSION: Our present findings provide a feasible strategy to extensively identify gene- and cell-specific enhancers from DHSs obtained by high-throughput sequencing, which will help reveal the transcriptional regulation and biological functions of genes in some specific cells.

A deep learning model for early risk prediction of heart failure with preserved ejection fraction by DNA methylation profiles combined with clinical features.

BACKGROUND: Heart failure with preserved ejection fraction (HFpEF), affected collectively by genetic and environmental factors, is the common subtype of chronic heart failure. Although the available risk assessment methods for HFpEF have achieved some progress, they were based on clinical or genetic features alone. Here, we have developed a deep learning framework, HFmeRisk, using both 5 clinical features and 25 DNA methylation loci to predict the early risk of HFpEF in the Framingham Heart Study Cohort. RESULTS: The framework incorporates Least Absolute Shrinkage and Selection Operator and Extreme Gradient Boosting-based feature selection, as well as a Factorization-Machine based neural network-based recommender system. Model discrimination and calibration were assessed using the AUC and Hosmer-Lemeshow test. HFmeRisk, including 25 CpGs and 5 clinical features, have achieved the AUC of 0.90 (95% confidence interval 0.88-0.92) and Hosmer-Lemeshow statistic was 6.17 (P = 0.632), which outperformed models with clinical characteristics or DNA methylation levels alone, published chronic heart failure risk prediction models and other benchmark machine learning models. Out of them, the DNA methylation levels of two CpGs were significantly correlated with the paired transcriptome levels (R < -0.3, P < 0.05). Besides, DNA methylation locus in HFmeRisk were associated with intercellular signaling and interaction, amino acid metabolism, transport and activation and the clinical variables were all related with the mechanism of occurrence of HFpEF. Together, these findings give new evidence into the HFmeRisk model. CONCLUSION: Our study proposes an early risk assessment framework for HFpEF integrating both clinical and epigenetic features, providing a promising path for clinical decision making.

DeteX: A highly accurate software for detecting SNV and InDel in single and paired NGS data in cancer research.

Background: Genetic testing is becoming more and more accepted in the auxiliary diagnosis and treatment of tumors. Due to the different performance of the existing bioinformatics software and the different analysis results, the needs of clinical diagnosis and treatment cannot be met. To this end, we combined Bayesian classification model (BC) and fisher exact test (FET), and develop an efficient software DeteX to detect SNV and InDel mutations. It can detect the somatic mutations in tumor-normal paired samples as well as mutations in a single sample. Methods: Combination of Bayesian classification model (BC) and fisher exact test (FET). Results: We detected SNVs and InDels in 11 TCGA glioma samples, 28 clinically targeted capture samples and 2 NCCL-EQA standard samples with DeteX, VarDict, Mutect, VarScan and GatkSNV. The results show that, among the three groups of samples, DeteX has higher sensitivity and precision whether it detects SNVs or InDels than other callers and the F1 value of DeteX is the highest. Especially in the detection of substitution and complex mutations, only DeteX can accurately detect these two kinds of mutations. In terms of single-sample mutation detection, DeteX is much more sensitive than the HaplotypeCaller program in Gatk. In addition, although DeteX has higher mutation detection capabilities, its running time is only .609 of VarDict, which is .704 and .343 longer than VarScan and MuTect, respectively. Conclusion: In this study, we developed DeteX to detect SNV and InDel mutations in single and paired samples. DeteX has high sensitivity and precision especially in the detection of substitution and complex mutations. In summary, DeteX from NGS data is a good SNV and InDel caller.

A machine learning-assisted model for renal urate underexcretion with genetic and clinical variables among Chinese men with gout.

OBJECTIVES: The objective of this study was to develop and validate a prediction model for renal urate underexcretion (RUE) in male gout patients. METHODS: Men with gout enrolled from multicenter cohorts in China were analyzed as the development and validation data sets. The RUE phenotype was defined as fractional excretion of uric acid (FEUA) <5.5%. Candidate genetic and clinical features were screened by the least absolute shrinkage and selection operator (LASSO) with 10-fold cross-validation. Machine learning algorithms (stochastic gradient descent (SGD), logistic regression, support vector machine) were performed to construct a predictive classifier of RUE. Models were assessed by the area under the receiver operating characteristic curve (AUC) and the precision-recall curve (PRC). RESULTS: One thousand two hundred thirty-eight and two thousand twenty-three patients were enrolled as the development and validation cohorts, with 1220 and 754 randomly chosen patients genotyped, respectively. Rs3775948.GG of SLC2A9/GLUT9, rs504915.AA of NRXN2/URAT1, and 7 clinical features (age, hypertension, nephrolithiasis, blood glucose, serum urate, urea nitrogen, and creatinine) were generated by LASSO. Two additional SNP variants (rs2231142.GG of ABCG2 and rs11231463.GG of SLC22A9/OAT7) were selected based on their contributions to gout in the development cohort and their reported effects on renal urate handling. The optimized classifiers yielded AUCs of ~0.914 and PRCs of ~0.980 using these 11 variables. The SGD model was conducted in the validation cohort with an AUC of 0.899 and the PRC of 0.957. CONCLUSIONS: A prediction model for RUE composed of four SNPs and readily accessible clinical features was established with acceptable accuracy for men with gout.

Single-cell RNA sequencing reveals the cellular and molecular changes that contribute to the progression of lung adenocarcinoma.

Pure ground glass nodules (GGNs) and solid nodules (SNs) represent early and relatively late stages of lung adenocarcinoma (LUAD) in radiology, respectively. The cellular and molecular characteristics of pure GGNs and SNs have not been comprehensively elucidated. Additionally, the mechanism driving the progression of lung adenocarcinoma from pure GGN to SN in radiology is also elusive. In this study, by analyzing the single-cell transcriptomic profiles of 76,762 cells from four pure GGNs, four SNs, and four normal tissues, we found that anti-tumor immunity mediated by NK and CD8+T cells gradually weakened with the progression of LUAD and humoral immunity mediated by plasma B cells was more active in SNs. Additionally, the proliferation ability of some special epithelial cell increased during the progression process from pure GGN to SN. Furthermore, stromal cells and M2 macrophages could assist the progression of LUAD. Through comprehensive analyses, we revealed dynamic changes in cellular components and intercellular interactions during the progression of LUAD. These findings could facilitate our understanding of LUAD and discovery of novel therapeutic targets.

Single-cell transcriptomic analysis of the tumor ecosystem of adenoid cystic carcinoma.

Adenoid cystic carcinoma (ACC) is a malignant tumor that originates from exocrine gland epithelial cells. We profiled the transcriptomes of 49,948 cells from paracarcinoma and carcinoma tissues of three patients using single-cell RNA sequencing. Three main types of the epithelial cells were identified into myoepithelial-like cells, intercalated duct-like cells, and duct-like cells by marker genes. And part of intercalated duct-like cells with special copy number variations which altered with MYB family gene and EN1 transcriptomes were identified as premalignant cells. Developmental pseudo-time analysis showed that the premalignant cells eventually transformed into malignant cells. Furthermore, MYB and MYBL1 were found to belong to two different gene modules and were expressed in a mutually exclusive manner. The two gene modules drove ACC progression into different directions. Our findings provide novel evidence to explain the high recurrence rate of ACC and its characteristic biological behavior.