Biology subjects

Yao, B.

Publications and source records attributed to Yao, B..

At least 37 records · Page 2Linked to original sources

bioRxiv2024

Deep5hmC: Predicting genome-wide 5-Hydroxymethylcytosine landscape via a multimodal deep learning model

5-hydroxymethylcytosine (5hmC), a critical epigenetic mark with a significant role in regulating tissue-specific gene expression, is essential for understanding the dynamic functions of the human genome. Using tissue-specific 5hmC sequencing data, we introduce Deep5hmC, a multimodal deep learning framework that integrates both the DNA sequence and the histone modification information to predict genome-wide 5hmC modification. The multimodal design of Deep5hmC demonstrates remarkable improvement in predicting both qualitative and quantitative 5hmC modification compared to unimodal versions of Deep5hmC and state-of-the-art machine learning methods. This improvement is demonstrated through benchmarking on a comprehensive set of 5hmC sequencing data collected at four time points during forebrain organoid development and across 17 human tissues. Notably, Deep5hmC showcases its practical utility by accurately predicting gene expression and identifying differentially hydroxymethylated regions in a case-control study of Alzheimers disease.

bioinformatics↗

bioRxiv2024

Characterization of the SARS-CoV-2 BA.5 Variants in H11-K18-hACE2 Hamsters

This study aims to comprehensively characterize the SARS-CoV-2 BA.5 variants using K18 hACE2 transgenic mice and golden hamsters as model organisms. Previous research on SARS-CoV-2 has utilized both mouse and hamster models, leading to conflicting results concerning the viruss lethality. In our study, the finding suggests that H11-K18 hACE2 golden hamsters closely mimic the disease progression observed in human COVID-19 cases caused by BA.5 variants, demonstrating consistent severity and symptoms comparable to severe infections. Additionally, hamsters exhibit heightened respiratory viral replication, accurately reflecting the clinical viral kinetics observed in humans. The study emphasizes the critical importance of selecting an appropriate animal model for SARS-CoV-2 research, while also providing robust support for the hypothesis that BA.5 variants contribute to fatal outcomes in COVID-19 cases. These findings highlight the pivotal role of the golden hamster model in advancing our understanding of the pathogenic mechanisms underlying SARS-CoV-2 variants, as well as in the development of targeted therapeutic strategies. Significance StatementOur research work explores groundbreaking insights that could reshape our understanding of COVID-19 and pave the way for targeted therapies. We use golden hamster models to express the possibility of different animal models could contribute to human diseases. We hope this finding could clarify some conflicts existed, and help further development of medication for COVID.

immunology↗

bioRxiv2024

scaDA: A Novel Statistical Method for Differential Analysis of Single-Cell Chromatin Accessibility Sequencing Data

Single-cell ATAC-seq sequencing data (scATAC-seq) has been widely used to investigate chromatin accessibility on the single-cell level. One important application of scATAC-seq data analysis is differential chromatin accessibility analysis. However, the data characteristics of scATAC-seq such as excessive zeros and large variability of chromatin accessibility across cells impose a unique challenge for DA analysis. Existing statistical methods focus on detecting the mean difference of the chromatin accessible regions while overlooking the distribution difference. Motivated by real data exploration that distribution difference exists among cell types, we introduce a novel composite statistical test named "scaDA", which is based on zero-inflated negative binomial model (ZINB), for performing differential distribution analysis of chromatin accessibility by jointly testing the abundance, prevalence and dispersion simultaneously. Benefiting from both dispersion shrinkage and iterative refinement of mean and prevalence parameter estimates, scaDA demonstrates its superiority to both ZINB-based likelihood ratio tests and published methods by achieving the highest power and best FDR control in a comprehensive simulation study. In addition to demonstrating the highest power in three real sc-multiome data analyses, scaDA successfully identifies differentially accessible regions in microglia from sc-multiome data for an Alzheimer s disease (AD) study, regions which are most enriched in GO terms related to neurogenesis, the clinical phenotype of AD, and SNPs identified in AD-associated GWAS. Author summaryUnderstanding the cis-regulatory elements that control the fundamental gene regulatory process is important to basic biology. scATAC-seq data offers an unprecedented opportunity to investigate chromatin accessibility on the single-cell level and explore cell heterogeneity to reveal the dynamic changes of cis-regulatory elements among different cell types. To understand the dynamic change of gene regulation using scATAC-seq data, differential chromatin (DA) analysis, which is one of the most fundamental analyses for scATAC-seq data, can enable the identification of differentially accessible regions between cell types or between multiple conditions. Subsequently, DA analysis has many applications such as identifying cell type-specific chromatin accessible regions to reveal the cell type-specific gene regulatory program, assessing disease-associated changes in chromatin accessibility to detect potential biomarkers, and linking differentially accessible regions to differentially expressed genes for building a comprehensive gene regulatory map. This paper proposes a novel statistical method named "scaDA" to improve the detection of differentially accessible regions by performing differential distribution analysis. scaDA is believed to benefit the research community of single-cell genomics.

bioinformatics↗

bioRxiv2023

HELZ promotes R loop resolution to facilitate DNA double-strand break repair by homologous recombination

R loop homeostasis is critical for DNA double-strand break (DSB) repair; however, how R loops are resolved in this context is poorly understood. Here, we define HELZ as a unique RNA-DNA helicase that resolves R loops to facilitate homologous recombination (HR) repair. From a synthetic lethal etoposide resistance siRNA screen, we found that HELZ depletion causes R loop-mediated hypersensitivity to DSB-inducing agents, and HELZ localizes and binds to DSBs. HELZ preferentially binds to and unwinds RNA-DNA hybrids with 5ssRNA overhangs to promote R loop resolution genome-wide and at DSBs. Interestingly, HELZ facilitates BRCA1 recruitment to DSBs by preventing R loop accumulation, thereby promoting DNA end resection and HR to prevent R loop mediated genomic instability. Our findings define a role for HELZ in resolving R loops critical for HR that promotes genome stability and governs DSB-inducing agent resistance.

molecular biology↗

bioRxiv2023

STASCAN deciphers fine-resolution cell-distribution maps in spatial transcriptomics by deep learning

BackgroundThe spatial transcriptomics (ST) technologies have been widely applied to decode the spatial distribution of cells by resolving gene expression profiles in tissues. However, a fine-resolved spatial cell map is still limited by algorithmic tools and sequencing techniques. ResultsHere we develop a novel deep learning approach, STASCAN, which could define the spatial cellular distribution of both captured and uncharted areas by cell feature learning that combines gene expression profiles and histology images. STASCAN additionally adopts optional transfer learning and pseudo-labeling methods to improve the accuracy of the cell-type prediction from images. We have successfully applied STASCAN to enhance cell resolution, and revealed finer organizational structures across diverse datasets from various species and tissues generated from 10x Visium technology. STASCAN improves cell resolution of Schmidtea mediterranea datasets by six times and reconstructs more detailed 3D cell-type models. Furthermore, STASCAN could accurately pinpoint the boundaries of distinct cell layers in human intestinal tissue, specifically identify a micrometer-scale smooth muscle bundle structure in consistent with anatomic insights in human lung tissue, and redraw the spatial structural variation with enhanced cell patterns in human myocardial infarction tissue. Additionally, through STASCAN on embryonic mouse brain datasets generated by DBiT-derived MISAR-seq technology, the increased cellular resolution and distinct anatomical tissue domains with cell-type niches are revealed. Collectively, STASCAN is compatible with different ST technologies and has notable advantages in generating cell maps solely from histology images, thereby enhancing the spatial cellular resolution. ConclusionsIn short, STASCAN displays significant advantages in deciphering higher-resolution cellular distribution, resolving enhanced organizational structures and demonstrating its potential applications in exploring cell-cell interactions within the tissue microenvironment.

bioinformatics↗

bioRxiv2023

Relationship between reactive astrocytes, by SMBT-1 imaging, with amyloid-beta, tau, glucose metabolism, and microgliosis in mouse models of Alzheimer's disease

PurposeReactive astrocytes play an important role in the development of Alzheimers disease (AD). Here, we aim to investigate the temporospatial relationship between reactive astrocytes, tau and amyloid-{beta}, glucose metabolism, and microgliosis by using multitracer imaging in AD transgenic mouse models. MethodsPositron emission tomography (PET) imaging with [18F]SMBT-1 (monoamine oxidase-B), [18F]florbetapir (A{beta}), [18F]PM-PBB3 (tau), [18F]FDG, and [18F]DPA-714 (translocator protein) was carried out in 5- and 10-month-old APP/PS1, 11-month-old 3xTg mice, and aged-matched wild-type mice. The brain regional referenced standard uptake value (SUVR) was computed with the cerebellum as the reference region. Immunofluorescence staining was performed in mouse brain tissue slices. Results[18F]SMBT-1 and [18F]florbetapir SUVRs were higher in the cortex and hippocampus of 10-month-old APP/PS1 mice than in 5-month-old APP/PS1 mice and wild-type mice. Reduced [18F]FDG SUVR was observed in the thalamus and midbrain of 5-month-old APP/PS1 mice compared to wild-type mice. No significant difference in brain regional [18F]DPA-714 SUVR was observed in 5- and 10-month-old APP/PS1 mice compared to wild- type mice. No significant difference in the SUVRs of any tracers was observed in 11-month-old 3xTg mice compared to age-matched wild-type mice. A positive correlation between the SUVRs of [18F]SMBT-1 and [18F]DPA-714 in the cortex was observed. Immunostaining validated the distribution of MAO-B and TSPO, amyloid and tau inclusions in brain tissue from 10-month-old APP/PS1 mice and limited changes in 11-month- old 3xTg mice. ConclusionThe findings provide in vivo evidence for reactive astrocytes along with amyloid plaque and tau deposition preceding microgliosis in animal models of AD pathologies.

neuroscience↗

bioRxiv2023

The RNA helicase DDX1 associates with the nuclear RNA exosome and modulates R-loops

The RNA exosome is a ribonuclease complex that mediates both RNA processing and degradation. This complex is evolutionarily conserved, ubiquitously expressed, and required for fundamental cellular functions, including rRNA processing. The RNA exosome plays roles in regulating gene expression and protecting the genome, including modulating the accumulation of RNA-DNA hybrids (R-loops). The function of the RNA exosome is facilitated by cofactors, such as the RNA helicase MTR4, which binds/remodels RNAs. Recently, missense mutations in RNA exosome subunit genes have been linked to neurological diseases. One possibility to explain why missense mutations in genes encoding RNA exosome subunits lead to neurological diseases is that the complex may interact with cell- or tissue-specific cofactors that are impacted by these changes. To begin addressing this question, we performed immunoprecipitation of the RNA exosome subunit, EXOSC3, in a neuronal cell line (N2A) followed by proteomic analyses to identify novel interactors. We identified the putative RNA helicase, DDX1, as an interactor. DDX1 plays roles in double-strand break repair, rRNA processing, and R-loop modulation. To explore the functional connections between EXOSC3 and DDX1, we examined the interaction following double-strand breaks, and analyzed changes in R-loops in N2A cells depleted for EXOSC3 or DDX1 by DNA/RNA immunoprecipitation followed by sequencing (DRIP-Seq). We find that EXOSC3 interaction with DDX1 is decreased in the presence of DNA damage and that loss of EXOSC3 or DDX1 alters R-loops. These results suggest EXOSC3 and DDX1 interact during events of cellular homeostasis and potentially suppress unscrupulous expression of genes promoting neuronal projection.

biochemistry↗

bioRxiv2023

Spatial Transcriptomics Arena (STAr): an Integrated Platform for Spatial Transcriptomics Methodology Research

The emerging field of spatially resolved transcriptomics (SRT) has revolutionized biomedical research. SRT quantifies expression levels at different spatial locations, providing a new and powerful tool to interrogate novel biological insights. An essential question in the analysis of SRT data is to identify spatially variable (SV) genes; the expression levels of such genes have spatial variation across different tissues. SV genes usually play an important role in underlying biological mechanisms and tissue heterogeneity. Currently, several computational methods have been developed to detect such genes; however, there is a lack of unbiased assessment of these approaches to guide researchers in selecting the appropriate methods for their specific biomedical applications. In addition, it is difficult for researchers to implement different existing methods for either biological study or methodology development. Furthermore, currently available public SRT datasets are scattered across different websites and preprocessed in different ways, posing additional obstacles for quantitative researchers developing computational methods for SRT data analysis. To address these challenges, we designed Spatial Transcriptomics Arena (STAr), an open platform comprising 193 curated datasets from seven technologies, seven statistical methods, and analysis results. This resource allows users to retrieve high-quality datasets, apply or develop spatial gene detection methods, as well as browse and compare spatial gene analysis results. It also enables researchers to comprehensively evaluate SRT methodology research in both simulated and real datasets. Altogether, STAr is an integrated research resource intended to promote reproducible research and accelerate rigorous methodology development, which can eventually lead to an improved understanding of biological processes and diseases. STAr can be accessed at https://lce.biohpc.swmed.edu/star/.

genomics↗

bioRxiv2022

Transcranial electrical stimulation modulates emotional experience and metabolites in the prefrontal cortex in a donation task

In this study we investigated whether transcranial Direct Current Stimulation (tDCS) applied to two prefrontal cortex regions, the ventromedial prefrontal cortex (vmPFC, anode) and right dorsolateral prefrontal cortex (DLPFC, cathode) can be used to induce changes in self-reported emotions and modulate metabolite concentrations under both electrodes. We employed in vivo quantitative MR Spectroscopy (MRS) in healthy adult participants and quantified the changes in GABA and Glx complex (Glutamate and Glutamine) before and after five sessions of tDCS stimulation. tDCS was delivered at 2mA intensity for 20 minutes for the active group and 1 minute for the sham group. tDCS was applied over five days while participants were engaged in a charitable donation task, known to engage the reward network. We observed increased levels of GABA in vmPFC, but not in DLPFC. Glx levels were decreased in both vmPFC and DLPFC. We found that self-reported happiness increased significantly over time only in the active group. There was a decrease in self-reported guiltiness in both active and sham groups. Altogether, the results indicate that self-reported emotions can be modulated by prefrontal stimulation. This may be because the changes in GABA and Glx concentrations following repeated stimulation induce remote changes in the reward network through interactions with other metabolites, previously thought to be unreachable with noninvasive stimulation techniques.

neuroscience↗

bioRxiv2022

Biological responses to terahertz radiation with different power density in primary hippocampal neurons

BackgroundTerahertz (THz) radiation is a valuable imaging and sensing tool which is widely used in industry and medicine. However, it biological effects including genotoxicity and cytotoxicity are not clear, particularly on the nervous system. In this study, we investigated how THz radiation with different frequency and intensity would affect the morphology, cell growth and function of rat hippocampal neurons in vitro. MethodsPrimary hippocampal neuronal cells were isolated from newborn Wistar rat pups and were exposed to THz radiation with the frequencies of 0.12 THz and 0.141 THz, at an power intensities of 10 mW/cm2 and 30 mW/cm2 respectively. The cellular activities were evaluated by Cell Counting Kit-8 after exposure for 10 min and 30 min. The cellular apoptosis was examined by Annexin V staining using flow cytometry and the ultrastructure was detected by electron microscopy. Moreover, the amount of amino acid in the cultured neurons was measured by high performance liquid chromatography (HPLC). ResultsNo obvious morphological changes, such as axon length and axon number, could be detected in primary neurons after exposure to 10mW/cm2 or 50 mW/cm2 THz for 30 min. However, the overall cellular activities were differentially affected by 10mW/cm2 and 50 mW/cm2 THz. Obvious cellular apoptosis was induced by both 10mW/cm2 and 50 mW/cm2 exposure. We also found that the amounts of most amino acids did not change significantly after exposure to these two types of THz radiation. ConclusionsThe effects of THz radiation on hippocampal neurons depend on frequency, power, exposure time as well as the measurement time after the exposure. These results will provide important information for further investigation of the effect of THz radiation on the nervous system.

neuroscience↗

bioRxiv2022

Genomic insights into the phylogeny and biomass-degrading enzymes of rumen ciliates

Understanding the biodiversity and genetics of the gut microbiome has important implications for host physiology. One underexplored and elusive group is ciliated protozoa, which play crucial roles in regulating gut microbial interactions. Integrating single-cell sequencing and an assembly-and-identification pipeline, we acquired 52 high-quality ciliate genomes of 22 rumen morphospecies for all major abundant clades. With these genomes, we firstly resolved the taxonomic and phylogenetic framework that reclassified them into 19 species spanning 13 genera and reassigned the genus Dasytricha from Isotrichidae to a new family Dasytrichidae. Via extensive horizontal gene transfer and gene family expansion, rumen ciliates possess a broad array of enzymes to synergistically degrade plant and microbial carbohydrates. In particular, [~]80% of the degrading enzymes in Diplodiniinae and Ophryoscolecinae act on plant cell wall, and the high activities of their cellulase, xylanase and lysozyme reflect the potential of ciliate enzymes for biomass-conversion. Additionally, the new ciliate dataset greatly facilitated the rumen metagenomic analyses by allowing [~]12% of reads to be classified.

microbiology↗

bioRxiv2021

Ratiometric 4Pi single molecule localization with optimal resolution and color assignment

4Pi single molecule localization microscopy (4Pi-SMLM) with two opposing objectives achieves sub-10 nm isotropic 3D resolution with as few as 250 photons collected by each objective. Here, we developed a new ratiometric multi-color imaging strategy for 4Pi-SMLM which employed the intrinsic multi-phase interference intensity without increasing the complexity of the system and achieved both optimal 3D resolution and color separation. By partially linking the photon parameters between channels with interference difference of {pi} during global fitting of the multi-channel 4Pi single molecule data, we showed on simulated data that the loss of the localization precision is minimal compared with the theoretical minimum uncertainty, the Cramer-Rao lower bound (CRLB).

biophysics↗

bioRxiv2021

Nanopore callers for epigenetics from limited supervised data

Nanopore sequencing platforms combined with supervised machine learning (ML) have been effective at detecting base modifications in DNA such as 5mC and 6mA. These ML-based nanopore callers have typically been trained on data that span all modifications on all possible DNA k-mer backgrounds--a complete training dataset. However, as nanopore technology is pushed to more and more epigenetic modifications, such complete training data will not be feasible to obtain. Nanopore calling has historically been performed with Hidden Markov Models (HMMs) that cannot make successful calls for k-mer contexts not seen during training because of their independent emission distributions. However, deep neural networks (DNNs), which share parameters across contexts, are increasingly being used as callers, often outperforming their HMM cousins. It stands to reason that a DNN approach should be able to better generalize to unseen k-mer contexts. Indeed, herein we demonstrate that a common DNN approach (DeepSignal) outperforms a common HMM approach (Nanopolish) in the incomplete data setting. Furthermore, we propose a novel hybrid HMM-DNN approach, Amortized-HMM, that outperforms both the pure HMM and DNN approaches on 5mC calling when the training data are incomplete. Such an approach is expected to be useful for calling 5hmC and combinations of cytosine modifications, where complete training data are not likely to be available.

bioinformatics↗

bioRxiv2020

WEVar: a novel statistical learning framework for predicting noncoding regulatory variants

Understanding the functional consequence of noncoding variants is of great interest. Though genome-wide association studies (GWAS) or quantitative trait locus (QTL) analyses have identified variants associated with traits or molecular phenotypes, most of them are located in the noncoding regions, making the identification of causal variants a particular challenge. Existing computational approaches developed for for prioritizing non-coding variants produce inconsistent and even conflicting results. To address these challenges, we propose a novel statistical learning framework, which directly integrates the precomputed functional scores from representative scoring methods. It will maximize the usage of integrated methods by automatically learning the relative contribution of each method and produce an ensemble score as the final prediction. The framework consists of two modes. The first "context-free" mode is trained using curated causal regulatory variants from a wide range of context and is applicable to predict noncoding variants of unknown and diverse context. The second "context-dependent" mode further improves the prediction when the training and testing variants are from the same context. By evaluating the framework via both simulation and empirical studies, we demonstrate that it outperforms integrated scoring methods and the ensemble score successfully prioritizes experimentally validated regulatory variants in multiple risk loci.

bioinformatics↗

bioRxiv2020

PD-L1 lncRNA splice promotes lung adenocarcinoma progression via enhancing c-Myc activity

Although blockade of programmed death-ligand 1 (PD-L1) to enhance T cell immune responses shows great promise in tumor immunotherapy, the efficacy of such immune-checkpoint inhibition strategy is limited for patients with solid tumors. The mechanism underlying the limited efficacy of PD-L1 inhibitors remains unclear. Here, we show that human lung adenocarcinoma, regardless of PD-L1 protein positive or negative, all produce a long non-coding RNA isoform of PD-L1 (PD-L1-lnc) via alternative splicing, which promotes lung adenocarcinoma proliferation and metastasis. PD-L1-lnc in various lung adenocarcinoma cells is significantly upregulated by IFN{gamma} in a manner similar to PD-L1 mRNA. Both in vitro and in vivo studies demonstrate that PD-L1-lnc increases proliferation and invasion but decreases apoptosis of lung adenocarcinoma cells. Mechanistically, PD-L1-lnc directly binds to c-Myc and enhances c-Myc transcriptional activity downstream in lung adenocarcinoma cells. Our results provide targeting PD-L1-lnc-c-Myc axis as a novel strategy for lung cancer therapy.

cancer biology↗

bioRxiv2020

Phase response analyses support a relaxation oscillator model of locomotor rhythm generation in Caenorhabditis elegans

Neural circuits coordinate with muscles and sensory feedback to generate motor behaviors appropriate to an animals environment. In C. elegans, the mechanisms by which the motor circuit generates undulations and modulates them based on the environment are largely unclear. We quantitatively analyzed C. elegans locomotion during free movement and during transient optogenetic muscle inhibition. Undulatory movements were highly asymmetrical with respect to the duration of bending and unbending during each cycle. Phase response curves induced by brief optogenetic inhibition of head muscles showed gradual increases and rapid decreases as a function of phase at which the perturbation was applied. A relaxation oscillator model based on proprioceptive thresholds that switch the active muscle moment was developed and is shown to quantitatively agree with data from free movement, phase responses, and previous results for gait adaptation to mechanical loadings. Our results suggest a neuromuscular mechanism underlying C. elegans motor pattern generation within a compact circuit.

neuroscience↗

bioRxiv2020

HARMONIES: A Hybrid Approach for Microbiome Networks Inference via Exploiting Sparsity

The human microbiome is a collection of microorganisms. They form complex communities and collectively affect host health. Recently, the advances in next-generation sequencing technology enable the high-throughput profiling of the human microbiome. This calls for a statistical model to construct microbial networks from the microbiome sequencing count data. As microbiome count data are high-dimensional and suffer from uneven sampling depth, over-dispersion, and zero-inflation, these characteristics can bias the network estimation and require specialized analytical tools. Here we propose a general framework, HARMONIES, a Hybrid Approach foR MicrobiOme Network Inferences via Exploiting Sparsity, to infer a sparse microbiome network. HARMONIES first utilizes a zero-inflated negative binomial (ZINB) distribution to model the skewness and excess zeros in the microbiome data, as well as incorporates a stochastic process prior for sample-wise normalization. This approach infers a sparse and stable network by imposing non-trivial regularizations based on the Gaussian graphical model. In comprehensive simulation studies, HARMONIES outperformed four other commonly used methods. When using published microbiome data from a colorectal cancer study, it discovered a novel community with disease-enriched bacteria. In summary, HARMONIES is a novel and useful statistical framework for microbiome network inference, and it is available at https://github.com/shuangj00/HARMONIES.

bioinformatics↗

bioRxiv2019

Computational staining of pathology images to study tumor microenvironment in lung cancer

The spatial organization of different types of cells in tumor tissues reveals important information about the tumor microenvironment (TME). In order to facilitate the study of cellular spatial organization and interactions, we developed a comprehensive nuclei segmentation and classification tool to characterize the TME from standard Hematoxylin and Eosin (H&E)-stained pathology images. This tool can computationally \"stain\" different types of cell nuclei in H&E pathology images to facilitate pathologists in analyzing the TME.\n\nA Mask Regional-Convolutional Neural Network (Mask-RCNN) model was developed to segment the nuclei of tumor, stromal, lymphocyte, macrophage, karyorrhexis and red blood cells in lung adenocarcinoma (ADC). Using this tool, we identified and classified cell nuclei and extracted 48 cell spatial organization-related features that characterize the TME. Using these features, we developed a prognostic model from the National Lung Screening Trial dataset, and independently validated the model in The Cancer Genome Atlas (TCGA) lung ADC dataset, in which the predicted high-risk group showed significantly worse survival than the low-risk group (pv= 0.001), with a hazard ratio of 2.23 [1.37-3.65] after adjusting for clinical variables. Furthermore, the image-derived TME features were significantly correlated with the gene expression of biological pathways. For example, transcription activation of both the T-cell receptor (TCR) and Programmed cell death protein 1 (PD1) pathways was positively correlated with the density of detected lymphocytes in tumor tissues, while expression of the extracellular matrix organization pathway was positively correlated with the density of stromal cells.\n\nThis study developed a deep learning-based analysis tool to dissect the TME from tumor tissue images. Using this tool, we demonstrated that the spatial organization of different cell types is predictive of patient survival and associated with the gene expression of biological pathways. Although developed from the pathology images of lung ADC, this model can be adapted into other types of cancers.

bioinformatics↗