Reduced-rank clustered coefficient regression for addressing multicollinearity in heterogeneous coefficient estimation.

Clustered coefficient regression (CCR) extends the classical regression model by allowing regression coefficients varying across observations and forming clusters of observations. It has become an increasingly useful tool for modeling the heterogeneous relationship between the predictor and response variables. A typical issue of existing CCR methods is that the estimation and clustering results can be unstable in the presence of multicollinearity. To address the instability issue, this paper introduces a low-rank structure of the CCR coefficient matrix and proposes a penalized non-convex optimization problem with an adaptive group fusion-type penalty tailor-made for this structure. An iterative algorithm is developed to solve this non-convex optimization problem with guaranteed convergence. An upper bound for the coefficient estimation error is also obtained to show the statistical property of the estimator. Empirical studies on both simulated datasets and a COVID-19 mortality rate dataset demonstrate the superiority of the proposed method to existing methods.

Successfully treated with siltuximab and prednisone in a 7-year-old girl with DOCK8-deficiency presenting as recurrent wart-like lesions: a case report.

Dedicator of cytokinesis 8 (DOCK8) deficiency represents a primary immunodeficiency with a wide range of clinical symptoms, including recurrent infections, atopy, and increased malignancy risk. This study presents a case of a 6-year-old girl with DOCK8 deficiency, characterized by severe, treatment-resistant herpetic infections who was successfully treated with siltuximab and glucocorticoids. The successful use of siltuximab in achieving remission highlights the pivotal role of interleukin-6 (IL-6) in DOCK8 deficiency pathogenesis and suggests that IL-6 modulation can be critical in managing DOCK8 deficiency-related viral infections, which may inform future therapeutic strategies for DOCK8 deficiency and similar immunodeficiencies.

Aggrephagy-related gene signature correlates with survival and tumor-associated macrophages in glioma: Insights from single-cell and bulk RNA sequencing.

BACKGROUND: Aggrephagy is a lysosome-dependent process that degrades misfolded protein condensates to maintain cancer cell homeostasis. Despite its importance in cellular protein quality control, the role of aggrephagy in glioma remains poorly understood. OBJECTIVE: To investigate the expression of aggrephagy-related genes (ARGs) in glioma and in different cell types of gliomas and to develop an ARGs-based prognostic signature to predict the prognosis, tumor microenvironment, and immunotherapy response of gliomas. METHODS: ARGs were identified by searching the Reactome database. We developed the ARGs-based prognostic signature (ARPS) using data from the Cancer Genome Atlas (TCGA, n = 669) by Lasso-Cox regression. We validated the robustness of the signature in clinical subgroups and CGGA cohorts (n = 970). Gene set enrichment analysis (GSEA) was used to identify the pathways enriched in ARPS subgroups. The correlations between ARGs and macrophages were also investigated at single cell level. RESULTS: A total of 44 ARGs showed heterogeneous expression among different cell types of gliomas. Five ARGs (HSF1, DYNC1H1, DYNLL2, TUBB6, TUBA1C) were identified to develop ARPS, an independent prognostic factor. GSEA showed gene sets of patients with high-ARPS were mostly enriched in cell cycle, DNA replication, and immune-related pathways. High-ARPS subgroup had higher immune cell infiltration states, particularly macrophages, Treg cells, and neutrophils. APRS had positive association with tumor mutation burden (TMB) and immunotherapy response predictors. At the single cell level, we found ARGs correlated with macrophage development and identified ARGs-mediated macrophage subtypes with distinct communication characteristics with tumor cells. VIM+ macrophages were identified as pro-inflammatory and had higher interactions with malignant cells. CONCLUSION: We identified a novel signature based on ARGs for predicting glioma prognosis, tumor microenvironment, and immunotherapy response. We highlight the ARGs-mediated macrophages in glioma exhibit classical features.

Application of a Novel Miniaturized Histopathologic Microscope for Ex Vivo Identifying Cerebral Glioma Margins Rapidly During Surgery: A Parallel Control Study.

PURPOSE: The purpose of our study is to assess the clinical performance of the DiveScope, a novel handheld histopathologic microscope in rapidly differentiating glioma from normal brain tissue during neurosurgery. METHODS: Thirty-two ex vivo specimens from 18 patients were included in the present study. The excised suspicious tissue was sequentially stained with sodium fluorescein and methylene blue and scanned with DiveScope during surgery. The adjacent tissue was sent to the department of pathology for frozen section examination. They would eventually be sent to the pathology department later for hematoxylin and eosin staining for final confirmation. The positive likelihood ratio, negative likelihood ratio, sensitivity, specificity, and area under the curve of the device were calculated. In addition, the difference in time usage between DiveScope and frozen sections was compared for the initial judgment. RESULTS: The sensitivity and specificity of the DiveScope after analyzing hematoxylin and eosin -staining sections, were 88.29% and 100%, respectively. In contrast, the sensitivity and specificity of the frozen sections histopathology were 100% and 75%, respectively. The area under the curve of the DiveScope and the frozen sections histopathology was not significant ( P =0.578). Concerning time usage, DiveScope is significantly much faster than the frozen sections histopathology no matter the size of tissue. CONCLUSION: Compared with traditional pathological frozen sections, DiveScope was faster and displayed an equal accuracy for judging tumor margins intraoperatively.

Protective Effects of Hinokitiol on Neuronal Ferroptosis by Activating the Keap1/Nrf2/HO-1 Pathway in Traumatic Brain Injury.

In this study, we investigated the effects of hinokitiol, a small-molecule natural compound, against neuronal ferroptosis after traumatic brain injury (TBI). A controlled cortical impact (CCI) mouse model and excess glutamate-treated HT-22 cells were used to study the effects of hinokitiol on TBI. Hinokitiol mitigated TBI brain tissue lesions and significantly improved neurological function. Neuron loss and iron deposition were ameliorated after hinokitiol administration. Hinokitiol alleviated excessive glutamate-induced intracellular reactive oxygen species (ROS), lipid peroxidation, and Fe2+ accumulation in HT-22. Mechanistically, hinokitiol upregulated heme oxygenase-1 (HO-1) expression, promoted nuclear factor-erythroid factor 2-related factor 2 (Nrf2) nuclear translocation, and inhibited the activation of microglia and astrocyte after TBI. These results suggest that hinokitiol has neuroprotective effects on rescuing cells from TBI-induced neuronal ferroptosis. In summary, hinokitiol is a potential therapeutic candidate for TBI by activating the Nrf2/Keap1/HO-1 signaling pathway.

The causal effects between selenium levels and the brain cortical structure: A two-sample Mendelian randomization study.

Extensive research has demonstrated the critical role of selenium (Se) and selenoproteins in brain function and cognition. However, the impact of Se on brain cortical structure remains enigmatic. Therefore, this study used Mendelian randomization (MR) analysis to investigate the causal effect between Se levels and brain cortical structure. METHODS: This study utilizes 11 genetic variants associated with Se level variations, extracted from a large-scale genome-wide association study (GWAS) encompassed circulating Se levels (n = 5477) and toenail Se levels (n = 4162) in the European population. Outcome data were sourced from the summary statistics of the ENIGMA Consortium, comprising GWAS data from 51,666 individuals. The variables include cortical surface area (SA), thickness (TH) at the global level, and 34 functional cortical regions evaluated by magnetic resonance imaging. The inverse-variance-weighted method was used as the primary estimate. Additionally, sensitivity analyses were conducted to detect potential violations of assumptions underlying MR. RESULTS: At the global level, Se levels were not correlated with SA but showed a significant negative correlation with TH (ß = -0.00485 mm, SE = 0.00192, p = .0115). Heterogeneity was observed across different brain regions, with positive correlations found between Se levels and the TH of the parahippocampal gyrus, superior frontal gyrus, and frontal pole, whereas negative correlations were found with the TH of the inferior parietal lobe and middle temporal lobe. Regarding SA, Se levels exhibit positive correlations with the pars triangularis, caudal anterior cingulate, inferior parietal lobe, and banks of the superior temporal sulcus. Conversely, negative correlations were observed with the medial orbitofrontal cortex, posterior cingulate gyrus, insula, and the middle, superior, and transverse gyrus of the temporal lobe. No pleiotropy was detected. RESULTS: This MR study indicated that Se levels causally influence the brain cortical structure.

Functional Bayesian networks for discovering causality from multivariate functional data.

Multivariate functional data arise in a wide range of applications. One fundamental task is to understand the causal relationships among these functional objects of interest. In this paper, we develop a novel Bayesian network (BN) model for multivariate functional data where conditional independencies and causal structure are encoded by a directed acyclic graph. Specifically, we allow the functional objects to deviate from Gaussian processes, which is the key to unique causal structure identification even when the functions are measured with noises. A fully Bayesian framework is designed to infer the functional BN model with natural uncertainty quantification through posterior summaries. Simulation studies and real data examples demonstrate the practical utility of the proposed model.

Operative treatment of cystic prolactinomas: a retrospective study.

BACKGROUND: The optimal therapeutic approach for cystic prolactinomas remains unclear. This study aimed to evaluate the remission rates of prolactinoma patients after surgical treatment and the risk factors affecting postoperative remission in cystic prolactinoma patients. METHODS: The clinical data were retrospectively compiled from 141 patients with prolactinomas (including 41 cases of cystic prolactinomas, 21 cases of solid microprolactinomas and 79 cases of solid macroprolactinomas) who underwent transsphenoidal surgery (TSS) between April 2013 and October 2021 at the First Affiliated Hospital of Sun Yat-sen University. RESULTS: Early postoperative remission was achieved in 65.83% (n = 27/41) of cystic prolactinomas, 80.95% (n = 17/21) of solid microprolactinomas and 40.51% (n = 32/79) of solid macroprolactinomas. The mean length of follow up in all patients was 43.95 ± 2.33 months (range: 6-105 months). The follow-up remission rates were 58.54%, 71.43% and 44.30% in cystic, solid micro- and solid macroprolactinomas, respectively. For cystic prolactinomas, the early postoperative remission rates in the patients with preoperative dopamine agonists (DA) treatment were significantly higher than those without preoperative DA treatment (p = 0.033), but the difference in the follow-up remission rates between these two groups was not significant (p = 0.209). Multivariate stepwise logistic regression analysis indicated that tumor size and preoperative prolactin (PRL) levels < 200 ng/ml were independent predictors for early postoperative remission in cystic prolactinomas. CONCLUSION: For cystic prolactinomas, tumor size and preoperative PRL levels were independent predictors of early postoperative remission. Preoperative DA therapy combined with TSS may be more beneficial to cystic prolactinoma patients.

Individualized causal discovery with latent trajectory embedded Bayesian networks.

Bayesian networks have been widely used to generate causal hypotheses from multivariate data. Despite their popularity, the vast majority of existing causal discovery approaches make the strong assumption of a (partially) homogeneous sampling scheme. However, such assumption can be seriously violated, causing significant biases when the underlying population is inherently heterogeneous. To this end, we propose a novel causal Bayesian network model, termed BN-LTE, that embeds heterogeneous samples onto a low-dimensional manifold and builds Bayesian networks conditional on the embedding. This new framework allows for more precise network inference by improving the estimation resolution from the population level to the observation level. Moreover, while causal Bayesian networks are in general not identifiable with purely observational, cross-sectional data due to Markov equivalence, with the blessing of causal effect heterogeneity, we prove that the proposed BN-LTE is uniquely identifiable under relatively mild assumptions. Through extensive experiments, we demonstrate the superior performance of BN-LTE in causal structure learning as well as inferring observation-specific gene regulatory networks from observational data.

The gut microbiota and metabolite profiles are altered in patients with spinal cord injury.

BACKGROUND: Metabolites secreted by the gut microbiota may play an essential role in microbiota-gut-central nervous system crosstalk. In this study, we explored the changes occurring in the gut microbiota and their metabolites in patients with spinal cord injury (SCI) and analyzed the correlations among them. METHODS: The structure and composition of the gut microbiota derived from fecal samples collected from patients with SCI (n = 11) and matched control individuals (n = 10) were evaluated using 16S rRNA gene sequencing. Additionally, an untargeted metabolomics approach was used to compare the serum metabolite profiles of both groups. Meanwhile, the association among serum metabolites, the gut microbiota, and clinical parameters (including injury duration and neurological grade) was also analyzed. Finally, metabolites with the potential for use in the treatment of SCI were identified based on the differential metabolite abundance analysis. RESULTS: The composition of the gut microbiota was different between patients with SCI and healthy controls. At the genus level, compared with the control group, the abundance of UBA1819, Anaerostignum, Eggerthella, and Enterococcus was significantly increased in the SCI group, whereas that of Faecalibacterium, Blautia, Escherichia-Shigella, Agathobacter, Collinsella, Dorea, Ruminococcus, Fusicatenibacter, and Eubacterium was decreased. Forty-one named metabolites displayed significant differential abundance between SCI patients and healthy controls, including 18 that were upregulated and 23 that were downregulated. Correlation analysis further indicated that the variation in gut microbiota abundance was associated with changes in serum metabolite levels, suggesting that gut dysbiosis is an important cause of metabolic disorders in SCI. Finally, gut dysbiosis and serum metabolite dysregulation was found to be associated with injury duration and severity of motor dysfunction after SCI. CONCLUSIONS: We present a comprehensive landscape of the gut microbiota and metabolite profiles in patients with SCI and provide evidence that their interaction plays a role in the pathogenesis of SCI. Furthermore, our findings suggested that uridine, hypoxanthine, PC(18:2/0:0), and kojic acid may be important therapeutic targets for the treatment of this condition.

Risk Factors for Cerebral Infarction After Microsurgical Clipping of Hunt-Hess Grade 0-2 Single Intracranial Aneurysm: A Retrospective Study.

OBJECTIVE: The study aimed to explore risk factors for cerebral infarction after microsurgical clipping in patients with Hunt-Hess grade 0-2 single intracranial aneurysms. METHODS: A total of 137 patients with Hunt-Hess grade 0-2 single intracranial aneurysms treated with microsurgical clipping between March 2017 and December 2020 were retrospectively enrolled. Patients were divided into 2 groups on the basis of the occurrence of cerebral infarction after surgery. RESULTS: Of 137 enrolled patients, 14 (10.22%) showed cerebral infarction symptoms after surgery. Univariate analysis indicated that ruptured aneurysm status, aneurysm rupture during surgery, history of transient ischemic attack (TIA)/stroke, aneurysm size ≥7 mm, temporary clipping, intraoperative systolic hypotension (IOH), and occurrences of intraoperative motor-evoked potentials change were significantly related to postoperative cerebral infarction (PCI). However, using multivariate regression, only history of TIA/stroke (odds ratio = 0.124; 95% confidence interval [CI] = 0.021-0.748, P = 0.023) and IOH (odds ratio = 0.032; 95% CI = 0.005-0.210, P < 0.001) were independent predictors for PCI. Receiver operating characteristic curve analysis showed that the critical duration of temporary clipping and IOH that minimized the risk of PCI was 5.5 minutes and 7.5 minutes, respectively. CONCLUSIONS: Our study identified history of TIA/stroke and IOH as independent risk factors for cerebral infarction after microsurgical clipping.

Transsphenoidal surgery for prolactinomas in male patients: a retrospective study.

Male patients with prolactinomas usually present with typical hyperprolactinemia symptoms, including sexual dysfunction and infertility. However, clinical factors related to sexual dysfunction and surgical outcomes in these patients remain unclear. This study aimed to investigate the outcomes of male patients with prolactinomas after transsphenoidal surgery and the risk factors affecting sexual dysfunction. This study was conducted on 58 male patients who underwent transsphenoidal surgery for prolactinomas between May 2014 and December 2020 at the First Affiliated Hospital of Sun Yat-sen University, Guangzhou, China. We evaluated the sexual function of patients before and after surgery through International Index of Erectile Function-5 scores, libido, and frequency of morning erection. Of the 58 patients, 48 (82.8%) patients had sexual intercourse preoperatively. Among those 48 patients, 41 (85.4%) patients presented with erectile dysfunction. The preoperative International Index of Erectile Function-5 scores in patients with macroprolactinomas were significantly higher than those in patients with giant prolactinomas (17.63 ± 0.91 vs 13.28 ± 1.43; P = 0.01). Postoperatively, the incidence of erectile dysfunction was 47.9%, which was significantly lower than that preoperatively (85.4%; P = 0.01). Twenty-eight (68.3%) patients demonstrated an improvement in erectile dysfunction. Tumor size and invasiveness were significantly correlated with the improvement of erectile dysfunction. Preoperative testosterone <2.3 ng ml-1 was an independent predictor of improvement in erectile dysfunction. In conclusion, our results indicated that tumor size and invasiveness were important factors affecting the improvement of sexual dysfunction in male patients with prolactinoma. The preoperative testosterone level was an independent predictor related to the improvement of erectile dysfunction.

A Unified Bayesian Framework for Bi-overlapping-Clustering Multi-omics Data via Sparse Matrix Factorization.

The advances of modern sequencing techniques have generated an unprecedented amount of multi-omics data which provide great opportunities to quantitatively explore functional genomes from different but complementary perspectives. However, distinct modalities/sequencing technologies generate diverse types of data which greatly complicate statistical modeling because uniquely optimized methods are required for handling each type of data. In this paper, we propose a unified framework for Bayesian nonparametric matrix factorization that infers overlapping bi-clusters for multi-omics data. The proposed method adaptively discretizes different types of observations into common latent states on which cluster structures are built hierarchically. The proposed Bayesian nonparametric method is able to automatically determine the number of clusters. We demonstrate the utility of the proposed method using simulation studies and applications to a single-cell RNA-sequencing dataset, a combination of single-cell RNA-sequencing and single-cell ATAC-sequencing dataset, a bulk RNA-sequencing dataset, and a DNA methylation dataset which reveal several interesting findings that are consistent with biological literature.

Circular EZH2-encoded EZH2-92aa mediates immune evasion in glioblastoma via inhibition of surface NKG2D ligands.

Glioblastoma (GBM) is a highly aggressive primary brain tumour and is resistant to nearly all available treatments, including natural killer (NK) cell immunotherapy. However, the factors mediating NK cell evasion in GBM remain largely unclear. Here, we report that EZH2-92aa, a protein encoded by circular EZH2, is overexpressed in GBM and induces the immune evasion of GBM stem cells (GSCs) from NK cells. Positively regulated by DEAD-box helicase 3 (DDX3), EZH2-92aa directly binds the major histocompatibility complex class I polypeptide-related sequence A/B (MICA/B) promoters and represses their transcription; it also indirectly represses UL16-binding protein (ULBP) transcription by stabilizing EZH2. The downregulation of NK group 2D ligands (NKG2DLs, including MICA/B and ULBPs) in GSCs mediates NK cell resistance. Moreover, stable EZH2-92aa knockdown enhances NK cell-mediated GSC eradication in vitro and in vivo and synergizes with anti-PD1 therapy. Our results highlight the immunosuppressive function of EZH2-92aa in inhibiting the NK cell response in GBM and the clinical potential of targeting EZH2-92aa for NK-cell-directed immune therapy.

LinDA: linear models for differential abundance analysis of microbiome compositional data.

Differential abundance analysis is at the core of statistical analysis of microbiome data. The compositional nature of microbiome sequencing data makes false positive control challenging. Here, we show that the compositional effects can be addressed by a simple, yet highly flexible and scalable, approach. The proposed method, LinDA, only requires fitting linear regression models on the centered log-ratio transformed data, and correcting the bias due to compositional effects. We show that LinDA enjoys asymptotic FDR control and can be extended to mixed-effect models for correlated microbiome data. Using simulations and real examples, we demonstrate the effectiveness of LinDA.

Bayesian biclustering for microbial metagenomic sequencing data via multinomial matrix factorization.

High-throughput sequencing technology provides unprecedented opportunities to quantitatively explore human gut microbiome and its relation to diseases. Microbiome data are compositional, sparse, noisy, and heterogeneous, which pose serious challenges for statistical modeling. We propose an identifiable Bayesian multinomial matrix factorization model to infer overlapping clusters on both microbes and hosts. The proposed method represents the observed over-dispersed zero-inflated count matrix as Dirichlet-multinomial mixtures on which latent cluster structures are built hierarchically. Under the Bayesian framework, the number of clusters is automatically determined and available information from a taxonomic rank tree of microbes is naturally incorporated, which greatly improves the interpretability of our findings. We demonstrate the utility of the proposed approach by comparing to alternative methods in simulations. An application to a human gut microbiome data set involving patients with inflammatory bowel disease reveals interesting clusters, which contain bacteria families Bacteroidaceae, Bifidobacteriaceae, Enterobacteriaceae, Fusobacteriaceae, Lachnospiraceae, Ruminococcaceae, Pasteurellaceae, and Porphyromonadaceae that are known to be related to the inflammatory bowel disease and its subtypes according to biological literature. Our findings can help generate potential hypotheses for future investigation of the heterogeneity of the human gut microbiome.

Treatment barriers and clinical outcome of children with medulloblastoma in China: a report from the Chinese Children's Cancer Group (CCCG).

BACKGROUND: Medulloblastoma (MB) is the most common malignant central nervous system tumor of childhood. Management requires interdisciplinary care and is associated with unique challenges in developing regions. Here, we report the characteristics, clinical outcome and treatment barriers for Chinese children with MB based on a multi-institutional cohort from the Chinese Children's Cancer Group (CCCG). METHODS: Retrospective cohort study among 12 Chinese pediatric oncology units from the CCCG Brain Tumor Workgroup on patients aged <18 years diagnosed with MB from 2016 to 2019. RESULTS: 221 patients (male:female = 138:83) were included, 175 (79%) were ≥3 years of age, and 46 (21%) <3 years. 177 patients (80%) were completely staged, among which 50 (28%) had metastasis and 70 (40%) were considered to have high-risk (HR) disease. Gross/near-total resection was achieved in 203 patients (92%). In patients where molecular grouping could be assigned, 19 (16%), 35 (29%), and 65 (54%), respectively had WNT-activated, SHH-activated, and Group 3/4 MB. The median duration between resection and initiation of adjuvant therapy was 36 days. Respective 2-year PFS and OS rates were 76.0 ± 3.0% and 88.0 ± 2.3%. PFS was significantly associated with age, metastatic status and clinical risk grouping. Chemotherapy use during CSI or alkylator choice were not significant predictors for patient outcome. CONCLUSIONS: We reported the clinical profiles and outcome from the largest cohort of Chinese children with MB after multi-modal therapy. Strengths and limitations on the local provision of neuro-oncology service are identified.

Assessing the Multivariate Relationship between the Human Infant Intestinal Exfoliated Cell Transcriptome (Exfoliome) and Microbiome in Response to Diet.

Gut microbiota and the host exist in a mutualistic relationship, with the functional composition of the microbiota strongly influencing the health and well-being of the host. In addition to the standard differential expression analysis of host genes to assess the complex cross-talk between environment (diet), microbiome, and host intestinal physiology, data-driven integrative approaches are needed to identify potential biomarkers of both host genes and microbial communities that characterize these interactions. Our findings demonstrate that the complementary application of univariate differential gene expression analysis and multivariate approaches such as sparse Canonical Correlation Analysis (sCCA) and sparse Principal Components Analysis (sPCA) can be used to integrate data from both the healthy infant gut microbial community and host transcriptome (exfoliome) using stool derived exfoliated cells shed from the gut. These approaches reveal host genes and microbial functional categories related to the feeding phenotype of the infants. Our findings also confirm that combinatorial noninvasive -omic approaches provide an integrative genomics-based perspective of neonatal host-gut microbiome interactions.

Challenges and opportunities for managing pediatric central nervous system tumors in China.

Central nervous system (CNS) tumors represent the most deadly cancer in pediatric age group. In China, thousands of children are diagnosed with CNS tumors every year. Despite the improving socioeconomic status and availability of medical expertise within the country, unique challenges remain for the delivery of pediatric neuro-oncology service. In this review, we discuss the existing hurdles for improving the outcome of children with CNS tumors in China. Need for precise disease burden estimation, lack of intra- and inter-hospital collaborative networks, high probability of treatment abandonment, along with financial toxicities from treatment represent the key challenges that Chinese healthcare providers encounter. The tremendous opportunities for advancing the status of pediatric neuro-oncology care in and beyond the country are explored.

Co-variability of the summer NDVIs on the eastern Tibetan Plateau and in the Lake Baikal region: Associated climate factors and atmospheric circulation.

The Tibetan Plateau and Siberia are both crucial regions in which the vegetation dynamics are sensitive to climate change. The variabilities in the normalized difference vegetation index (NDVI) over the two regions have been explored previously, but there have been few studies on the relationship of the NDVI in the two regions. Using the GIMMS-NDVI, GHCN-CAMS and NCEP reanalysis datasets and statistical and physical diagnostic methods, we show that the summer (June, July and August) NDVI over the eastern Tibetan Plateau and Lake Baikal and its adjacent eastern region of Siberia have an in-phase co-variability, especially on an interannual timescale (with a correlation coefficient of 0.69 during the time period 1982-2014). Further analyses show that precipitation and the related cloud cover and solar radiation are responsible for the variability in the NDVI over the eastern Tibetan Plateau, whereas temperature has the more important role in modulating the variability in the NDVI over the Lake Baikal region. A dipole pattern prevails over the Tibetan Plateau-Lake Baikal region and reflects the anomalies in the intensity and location of the South Asian high and the northeast Asian blocking high. This dipole pattern simultaneously modulates precipitation over the eastern Tibetan Plateau and the temperature over the Lake Baikal region and leads to the co-variability of the NDVI between the two regions. A synergistic sea surface temperature index, which reflects sea surface temperature anomalies in the eastern tropical Pacific Ocean, the northwest Pacific Ocean, the northern Indian Ocean and the subtropical north Atlantic Ocean, appears to adjust this Tibetan Plateau-Lake Baikal dipole pattern and is therefore closely related to the co-variability of the NDVI between the eastern Tibetan Plateau and the Lake Baikal region. Our results suggest that vegetation dynamics may not be only a local phenomenon in some areas, but are also likely to remotely link with variations in vegetation over other regions.