A review of the application of deep learning in obesity: From early prediction aid to advanced management assistance.

BACKGROUND AND AIMS: Obesity is a chronic disease which can cause severe metabolic disorders. Machine learning (ML) techniques, especially deep learning (DL), have proven to be useful in obesity research. However, there is a dearth of systematic reviews of DL applications in obesity. This article aims to summarize the current trend of DL usage in obesity research. METHODS: An extensive literature review was carried out across multiple databases, including PubMed, Embase, Web of Science, Scopus, and Medline, to collate relevant studies published from January 2018 to September 2023. The focus was on research detailing the application of DL in the context of obesity. We have distilled critical insights pertaining to the utilized learning models, encompassing aspects of their development, principal results, and foundational methodologies. RESULTS: Our analysis culminated in the synthesis of new knowledge regarding the application of DL in the context of obesity. Finally, 40 research articles were included. The final collection of these research can be divided into three categories: obesity prediction (n = 16); obesity management (n = 13); and body fat estimation (n = 11). CONCLUSIONS: This is the first review to examine DL applications in obesity. It reveals DL's superiority in obesity prediction over traditional ML methods, showing promise for multi-omics research. DL also innovates in obesity management through diet, fitness, and environmental analyses. Additionally, DL improves body fat estimation, offering affordable and precise monitoring tools. The study is registered with PROSPERO (ID: CRD42023475159).

Single-cell profiling identifies IL1Bhi macrophages associated with inflammation in PD-1 inhibitor-induced inflammatory arthritis.

Inflammatory arthritis (IA) is a common rheumatic adverse event following immune checkpoint inhibitors treatment. The clinical disparities between IA and rheumatoid arthritis (RA) imply disease heterogeneity and distinct mechanisms, which remain elusive. Here, we profile CD45+ cells from the peripheral blood or synovial fluid (SF) of patients with PD-1-induced IA (PD-1-IA) or RA using single-cell RNA sequencing. We report the predominant expansion of IL1Bhi myeloid cells with enhanced NLRP3 inflammasome activity, in both the SF and peripheral blood of PD-1-IA, but not RA. IL1Bhi macrophages in the SF of PD-1-IA shared similar inflammatory signatures and might originate from peripheral IL1Bhi monocytes. Exhausted CD8+ T cells (Texs) significantly accumulated in the SF of patients with PD-1-IA. IL1Bhi myeloid cells communicated with CD8+ Texs possibly via the CCR1-CCL5/CCL3 and CXCL10-CXCR3 axes. Collectively, these results demonstrate different cellular and molecular pathways in PD-1-IA and RA and highlight IL1Bhi macrophages as a possible therapeutic target in PD-1-IA.

Manganese Mineralization of Pathogenic Viruses as a Universal Vaccine Platform.

Biomimetic viral mineralization improves viral vaccine stability and immunogenicity using inorganic metals such as Ca, Al, or Fe. Mn is a metal found in high concentrations in mammalian tissues; however, under natural or laboratory conditions, Mn mineralization by medical viruses has yet to be established. Herein, a single IAV particle is successfully encapsulated with manganese phosphate (MnP) under specific conditions using the human influenza A virus (IAV). MnP-mineralized IAVs (IAV@Mn) exhibited physiochemical and in vitro properties similar to Ca-mineralized IAVs. In animal models, IAV@Mn shows limited replication in immune-competent cells and a significant attenuation compared to naïve cells. Moreover, a single-dose vaccination with IAV@Mn induced robust humoral and cellular immune responses and conferred significant protection against a wild-type IAV challenge in mice. Thus, Mn mineralization in pathogenic viruses provides a rapid and universal strategy for generating an emergency vaccine in response to emerging viruses.

MicroRNAs and their regulators: Potential therapeutic targets in pulmonary arterial hypertension.

Pulmonary arterial hypertension (PAH) is a complex and progressive disease characterized by pulmonary arterial remodeling. Despite that current combination therapy has shown improvement in morbidity and mortality, a better deciphering of the underlying pathological mechanisms and novel therapeutic targets is urgently needed to combat PAH. MicroRNA, the critical element in post-transcription mechanisms, mediates cellular functions mainly by tuning downstream target gene expression. Meanwhile, upstream regulators can regulate miRNAs in synthesis, transcription, and function. In vivo and in vitro studies have suggested that miRNAs and their regulators are involved in PAH. However, the miRNA-related regulatory mechanisms governing pulmonary vascular remodeling and right ventricular dysfunction remain elusive. Hence, this review summarized the controversial roles of miRNAs in PAH pathogenesis, focused on different miRNA-upstream regulators, including transcription factors, regulatory networks, and environmental stimuli, and finally proposed the prospects and challenges for the therapeutic application of miRNAs and their regulators in PAH treatment.

Identification of Anti-SNRPA as a Novel Serological Biomarker for Systemic Sclerosis Diagnosis.

Systemic sclerosis (SSc) is a systemic autoimmune disorder that leads to vasculopathy and tissue fibrosis. A lack of reliable biomarkers has been a challenge for clinical diagnosis of the disease. We employed a protein array-based approach to identify and validate SSc-specific autoantibodies. Phase I involved profiled autoimmunity using human proteome microarray (HuProt arrays) with 90 serum samples: 40 patients with SSc, 30 patients diagnosed with autoimmune diseases, and 20 healthy subjects. In Phase II, we constructed a focused array with candidates identified antigens and used this to profile a much larger cohort comprised of serum samples. Finally, we used a western blot analysis to validate the serum of validated proteins with high signal values. Bioinformatics analysis allowed us to identify 113 candidate autoantigens that were significantly associated with SSc. This two-phase strategy allowed us to identify and validate anti-small nuclear ribonucleoprotein polypeptide A (SNRPA) as a novel SSc-specific serological biomarker. The observed positive rate of anti-SNRPA antibody in patients with SSc was 11.25%, which was significantly higher than that of any disease control group (3.33%) or healthy controls (1%). In conclusion, anti-SNRPA autoantibody serves as a novel biomarker for SSc diagnosis and may be promising for clinical applications.

Single-cell analysis of peripheral blood from high-altitude pulmonary hypertension patients identifies a distinct monocyte phenotype.

Immune and inflammatory responses have an important function in the pathophysiology of pulmonary hypertension (PH). However, little is known about the immune landscape in peripheral circulation in patients with high-altitude pulmonary hypertension (HAPH). We apply single-cell transcriptomics to characterize the monocytes that are significantly enriched in the peripheral blood mononuclear cells (PBMC) of HAPH patients. We discover an increase in C1 (non-classical) and C2 (intermediate) monocytes in PBMCs and a decrease in hypoxia-inducible transcription factor-1α (HIF-1α) in all monocyte subsets associated with HAPH. In addition, we demonstrate that similar immune adaptations may exist in HAPH and PH. Overall, we characterize an immune cell atlas of the peripheral blood in HAPH patients. Our data provide evidence that specific monocyte subsets and HIF-1α downregulation might be implicated in the pathogenesis of HAPH.

Oncogene or tumor suppressor gene: An integrated pan-cancer analysis of NBPF1.

Neuroblastoma breakpoint family, member 1 (NBPF1), appears to be a double-edged sword with regard to its role in carcinogenesis. On the one hand, the tumor-suppressing functions of NBPF1 have been definitively observed in neuroblastoma, prostate cancer, cutaneous squamous cell carcinoma, and cervical cancer. On the other hand, there is evidence that NBPF1 regulates the colony formation, invasion, and maintenance of liver cancer cells and hence functions as an oncogene. The roles of NBPF1 are strictly dependent on the biological context and type of organization. However, a systematic pan-cancer analysis has thus far not been undertaken, and the significance of NBPF1 in the occurrence and progression of many malignancies is uncertain. In this paper, bioinformatics techniques were employed to analyze NBPF1 expression across different cancers and investigate the relationship between NBPF1 and clinical features, prognosis, genetic alteration, and tumor immune microenvironment, respectively. Our results show that NBPF1 is variably expressed in distinct tumor tissues and is also closely linked to clinical outcomes. In particular, compared to other tumor types, there was a strong negative correlation between NBPF1 expression and various components of the tumor microenvironment in adrenocortical carcinoma (ACC). We thus developed an NBPF1-derived immune risk model based on NBPF1-related immune genes; ACC patients with a high-risk score tended to have a poorer prognosis, accompanied by immune hyporesponsiveness. NBPF1 can be used as a prognostic biomarker for multiple cancers. Moreover, anti-NBPF1 immunotherapy may be suitable for treating ACC patients.

Exploration of Target Spaces in the Human Genome for Protein and Peptide Drugs.

After decades of development, protein and peptide drugs have now grown into a major drug class in the marketplace. Target identification and validation are crucial for the discovery of protein and peptide drugs, and bioinformatics prediction of targets based on the characteristics of known target proteins will help improve the efficiency and success rate of target selection. However, owing to the developmental history in the pharmaceutical industry, previous systematic exploration of the target spaces has mainly focused on traditional small-molecule drugs, while studies related to protein and peptide drugs are lacking. Here, we systematically explore the target spaces in the human genome specifically for protein and peptide drugs. Compared with other proteins, both successful protein and peptide drug targets have many special characteristics, and are also significantly different from those of small-molecule drugs in many aspects. Based on these features, we develop separate effective genome-wide target prediction models for protein and peptide drugs. Finally, a user-friendly web server, Predictor Of Protein and PeptIde drugs' therapeutic Targets (POPPIT) (http://poppit.ncpsb.org.cn/), is established, which provides not only target prediction specifically for protein and peptide drugs but also abundant annotations for predicted targets.

Identification and Validation of Immune Molecular Subtypes in Pancreatic Ductal Adenocarcinoma: Implications for Prognosis and Immunotherapy.

Background: Pancreatic ductal adenocarcinoma (PDAC) remains treatment refractory. Immunotherapy has achieved success in the treatment of multiple malignancies. However, the efficacy of immunotherapy in PDAC is limited by a lack of promising biomarkers. In this research, we aimed to identify robust immune molecular subtypes of PDAC to facilitate prognosis prediction and patient selection for immunotherapy. Methods: A training cohort of 149 PDAC samples from The Cancer Genome Atlas (TCGA) with mRNA expression data was analyzed. By means of non-negative matrix factorization (NMF), we virtually dissected the immune-related signals from bulk gene expression data. Detailed immunogenomic and survival analyses of the immune molecular subtypes were conducted to determine their biological and clinical relevance. Validation was performed in five independent datasets on a total of 615 samples. Results: Approximately 31% of PDAC samples (46/149) had higher immune cell infiltration, more active immune cytolytic activity, higher activation of the interferon pathway, a higher tumor mutational burden (TMB), and fewer copy number alterations (CNAs) than the other samples (all P < 0.001). This new molecular subtype was named Immune Class, which served as an independent favorable prognostic factor for overall survival (hazard ratio, 0.56; 95% confidence interval, 0.33-0.97). Immune Class in cooperation with previously reported tumor and stroma classifications had a cumulative effect on PDAC prognostic stratification. Moreover, programmed cell death-1 (PD-1) inhibitors showed potential efficacy for Immune Class (P = 0.04). The robustness of our immune molecular subtypes was further verified in the validation cohort. Conclusions: By capturing immune-related signals in the PDAC tumor microenvironment, we reveal a novel molecular subtype, Immune Class. Immune Class serves as an independent favorable prognostic factor for overall survival in PDAC patients.

Identification of the candidate lncRNA biomarkers for acute kidney injury: a systematic review and meta-analysis.

Background: This meta-analysis aims to summarize the studies of lncRNAs dysregulation in individual acute kidney injury (AKI) and identify the potential lncRNA biomarkers of AKI.Research design and methods: We systematically searched four databases to identify the lncRNA expression studies of AKI in animal models and patients. The lncRNAs expression data were extracted from 38 included studies, and lncRNA vote-counting strategy was applied to identify significant lncRNA biomarkers. The predicted targets of lncRNA biomarkers were obtained by searching Co-LncRNA, RBPmap, and LncBase v.2. Further, GO enrichment analysis and KEGG pathway analysis were performed.Results: We recognized a significant lncRNA signature of 21 up-regulated and 11 down-regulated lncRNAs, among which TapSAKI, XIST, MALAT1, CASC2, and HOXA-AS2 were dysregulated both in AKI rodent models and patients. About 28.0% of these lncRNAs mainly exist in the nucleus, which was also the most enriched GO cellular components term. The most relevant GO terms in biological process and molecular function associated with these lncRNAs were splicing, processing, and binding of mRNA.Conclusions: The present meta-analysis identified 31 significant dysregulated lncRNAs from 38 studies. TapSAKI, XIST, MALAT1, CASC2, and HOXA-AS2 were considered as the potential predictive biomarkers and therapeutic targets of AKI.

Generation of an induced pluripotent stem cell line (PUMCHi006-A) derived from a patient with pulmonary arterial hypertension carrying heterozygous c.1339 G > A mutation in PTGIS gene.

Our previous study found that mutations in the PTGIS gene contributed high susceptibility to pulmonary arterial hypertension (PAH). We have generated disease-specific induced pluripotent stem cell (iPSC) lines from a PAH patient carrying the heterozygous c.1339 G > A mutation in PTGIS gene. The generated iPSC lines can be differentiated into endothelial cells to investigate the pathogenesis of PAH associated with PTGIS gene, which could provide valuable resources for personalized medicine.

Next-generation sequencing reveals heterogeneous genetic alterations in key signaling pathways of mismatch repair deficient colorectal carcinomas.

Colorectal carcinoma (CRC) with deficient mismatch repair (dMMR) is an etiologically heterogeneous molecular entity. We investigated the genetic profile, focusing on key signaling pathways and molecular diversity of dMMR CRCs. In this study, next-generation sequencing was applied to 156 consecutive dMMR CRCs and 225 randomly selected proficient MMR (pMMR) CRCs diagnosed between July 2015 and December 2019 at Peking Union Medical College Hospital. Genetic alterations and MLH1 promoter hypermethylation (MLH1me+) were analyzed. Among the most frequently mutated genes, RNF43, ARID1A, PIK3CA, ATM, and BRCA2 mutants were enriched in dMMR CRCs, whereas APC and TP53 mutations were enriched in pMMR CRCs. In dMMR group, RNF43, APC, ARID1A, and BRCA2 mutations were largely microsatellite instability events. WNT pathway was commonly altered regardless of MMR status. Compared to pMMR CRCs, dMMR CRCs had remarkably more prevalent PI3K, RTK-RAS, TGFß, and DNA damage repair pathway alterations and more multiple mutations in WNT and PI3K pathways. Within dMMR tumors, mutual exclusivity occurred between CTNNB1 mutation and APC or RNF43 mutation, while coexistence existed between BRAF and RNF43 mutation, as well as RAS and APC mutation. MLH1me+ dMMR CRCs had significantly more frequent RNF43 mutations but less frequent KRAS, APC, and CTNNB1 mutations comparing to MLH1-unmethylated dMMR CRCs. RNF43/BRAF comutations were detected in MLH1me+ dMMR CRCs, whereas RAS/APC comutations were largely detected in Lynch syndrome-associated cases. RNF43 mutation was independently associated with MLH1me+ rather than BRAF mutations. dMMR CRCs bearing receptor tyrosine kinase fusion demonstrated no additional RTK-RAS mutations, significantly fewer PI3K alterations and more TGFBR2 mutations than other dMMR tumors. Our study revealed that dMMR CRCs had distinctive gene mutation spectra and signaling pathway interaction patterns compared to proficient mismatch repair (pMMR) CRCs, and molecular heterogeneity was evident for these divergent oncogenic pathways. These findings justify the use of individualized therapy targeted to dMMR CRC subgroups.

Discovery of Urinary Proteomic Signature for Differential Diagnosis of Acute Appendicitis.

Acute appendicitis is one of the most common acute abdomens, but the confident preoperative diagnosis is still a challenge. In order to profile noninvasive urinary biomarkers that could discriminate acute appendicitis from other acute abdomens, we carried out mass spectrometric experiments on urine samples from patients with different acute abdomens and evaluated diagnostic potential of urinary proteins with various machine-learning models. Firstly, outlier protein pools of acute appendicitis and controls were constructed using the discovery dataset (32 acute appendicitis and 41 control acute abdomens) against a reference set of 495 normal urine samples. Ten outlier proteins were then selected by feature selection algorithm and were applied in construction of machine-learning models using naïve Bayes, support vector machine, and random forest algorithms. The models were assessed in the discovery dataset by leave-one-out cross validation and were verified in the validation dataset (16 acute appendicitis and 45 control acute abdomens). Among the three models, random forest model achieved the best performance: the accuracy was 84.9% in the leave-one-out cross validation of discovery dataset and 83.6% (sensitivity: 81.2%, specificity: 84.4%) in the validation dataset. In conclusion, we developed a 10-protein diagnostic panel by the random forest model that was able to distinguish acute appendicitis from confusable acute abdomens with high specificity, which indicated the clinical application potential of noninvasive urinary markers in disease diagnosis.

Mutational profiling of poorly differentiated and anaplastic thyroid carcinoma by the use of targeted next-generation sequencing.

AIMS: To characterise the mutational profiles of poorly differentiated thyroid carcinoma (PDTC) and anaplastic thyroid carcinoma (ATC) and to identify markers with potential diagnostic, prognostic and therapeutic significance. METHODS AND RESULTS: Targeted next-generation sequencing with a panel of 18 thyroid carcinoma-related genes was performed on tissue samples from 41 PDTC and 25 ATC patients. Genetic alterations and their correlations with clinicopathological factors, including survival outcomes, were also analysed. Our results showed that ATC had significantly higher mutation rates of BRAF, TP53, TERT and PIK3CA than PDTC (P = 0.005, P = 0.007, P = 0.005, and P = 0.033, respectively). Nine (69%) ATC cases with papillary thyroid carcinoma (PTC) components harboured BRAF mutations, all of which coexisted with a late mutation event (TP53, TERT, or PIK3CA). Nine cases with oncogenic fusion (six RET cases, one NTRK1 case, one ALK case, and one PPARG case) were identified in 41 PDTCs, whereas only one case with oncogenic fusion (NTRK1) was found among 25 ATCs. Moreover, all six cases of RET fusion were found in PDTC with PTC components, accounting for 33%. In PDTC/ATC patients, concurrent TERT and PIK3CA mutations were associated with poor overall survival after adjustment for TNM stage (P = 0.001). CONCLUSIONS: ATC with PTC components is typically characterised by a BRAF mutation with a late mutation event, whereas PDTC with PTC components is more closely correlated with RET fusion. TERT and concurrent PIK3CA mutations predict worse overall survival in PDTC/ATC patients.

OsteoporosAtlas: a human osteoporosis-related gene database.

BACKGROUND: Osteoporosis is a common, complex disease of bone with a strong heritable component, characterized by low bone mineral density, microarchitectural deterioration of bone tissue and an increased risk of fracture. Due to limited drug selection for osteoporosis and increasing morbidity, mortality of osteoporotic fractures, osteoporosis has become a major health burden in aging societies. Current researches for identifying specific loci or genes involved in osteoporosis contribute to a greater understanding of the pathogenesis of osteoporosis and the development of better diagnosis, prevention and treatment strategies. However, little is known about how most causal genes work and interact to influence osteoporosis. Therefore, it is greatly significant to collect and analyze the studies involved in osteoporosis-related genes. Unfortunately, the information about all these osteoporosis-related genes is scattered in a large amount of extensive literature. Currently, there is no specialized database for easily accessing relevant information about osteoporosis-related genes and miRNAs. METHODS: We extracted data from literature abstracts in PubMed by text-mining and manual curation. Moreover, a local MySQL database containing all the data was developed with PHP on a Windows server. RESULTS: OsteoporosAtlas (http://biokb.ncpsb.org/osteoporosis/), the first specialized database for easily accessing relevant information such as osteoporosis-related genes and miRNAs, was constructed and served for researchers. OsteoporosAtlas enables users to retrieve, browse and download osteoporosis-related genes and miRNAs. Gene ontology and pathway analyses were integrated into OsteoporosAtlas. It currently includes 617 human encoding genes, 131 human non-coding miRNAs, and 128 functional roles. We think that OsteoporosAtlas will be an important bioinformatics resource to facilitate a better understanding of the pathogenesis of osteoporosis and developing better diagnosis, prevention and treatment strategies.

Revealing the metabolic characteristics of human embryonic stem cells by genome-scale metabolic modeling.

Embryonic stem cells (ESCs) are characterized by a dual capacity, self-renewal and pluripotency, which can be regulated by metabolism. A better understanding of ESC metabolism and regulatory mechanisms is pivotal for research into development, ageing, and cancer treatment. However, a systematic and comprehensive delineation of human ESC metabolism is still lacking. Here, we reconstructed the first genome-scale metabolic model (GEM) of human ESCs (hESCs). By GEM simulation and analyses, hESC global metabolic characteristics including essential metabolites and network motifs were identified. Potential metabolic subsystems responsible for self-renewal and pluripotency were also identified by analyses and experiments. This first GEM of hESCs provides a novel view and resource for stem cell metabolism research and will contribute to the elucidation of their metabolic characteristics.

Differential antiviral immunity to Japanese encephalitis virus in developing cortical organoids.

Japanese encephalitis (JE) caused by Japanese encephalitis virus (JEV) poses a serious threat to the world's public health yet without a cure. Certain JEV-infected neural cells express a subset of previously identified intrinsic antiviral interferon stimulated genes (ISGs), indicating brain cells retain autonomous antiviral immunity. However, whether this happens in composited brain remains unclear. Human pluripotent stem cell (hPSC)-derived organoids can model disorders caused by human endemic pathogens such as Zika virus, which may potentially address this question and facilitate the discovery of a cure for JE. We thus generated telencephalon organoid and infected them with JEV. We found JEV infection caused significant decline of cell proliferation and increase of cell death in brain organoid, resulting in smaller organoid spheres. JEV tended to infect astrocytes and neural progenitors, especially the population representing outer radial glial cells (oRGCs) of developing human brain. In addition, we revealed variable antiviral immunity in brain organoids of different stages of culture. In organoids of longer culture (older than 8 weeks), but not of early ones (less than 4 weeks), JEV infection caused typical activation of interferon signaling pathway. Preferential infection of oRGCs and differential antiviral response at various stages might explain the much more severe outcomes of JEV infection in the younger, which also provide clues to develop effective therapeutics of such diseases.

An integrated bioinformatics platform for investigating the human E3 ubiquitin ligase-substrate interaction network.

The ubiquitination mediated by ubiquitin activating enzyme (E1), ubiquitin conjugating enzyme (E2), and ubiquitin ligase (E3) cascade is crucial to protein degradation, transcription regulation, and cell signaling in eukaryotic cells. The high specificity of ubiquitination is regulated by the interaction between E3 ubiquitin ligases and their target substrates. Unfortunately, the landscape of human E3-substrate network has not been systematically uncovered. Therefore, there is an urgent need to develop a high-throughput and efficient strategy to identify the E3-substrate interaction. To address this challenge, we develop a computational model based on multiple types of heterogeneous biological evidence to investigate the human E3-substrate interactions. Furthermore, we provide UbiBrowser as an integrated bioinformatics platform to predict and present the proteome-wide human E3-substrate interaction network ( http://ubibrowser.ncpsb.org ).Protein stability modulation by E3 ubiquitin ligases is an important layer of functional regulation, but screening for E3 ligase-substrate interactions is time-consuming and costly. Here, the authors take an in silico naïve Bayesian classifier approach to integrate multiple lines of evidence for E3-substrate prediction, enabling prediction of the proteome-wide human E3 ligase interaction network.

Functional constraints on adaptive evolution of protein ubiquitination sites.

It is still unclear whether there exist functional constraints on the evolution of protein ubiquitination sites, because most previous studies regarded all protein ubiquitination sites as a whole or only focused on limited structural properties. We tried to clarify the relation between functional constraints and ubiquitination sites evolution. We investigated the evolutionary conservation of human ubiquitination sites in a broad evolutionary scale from G. gorilla to S. pombe, and we found that in organisms originated after the divergence of vertebrate, ubiquitination sites are more conserved than their flanking regions, while the opposite tendency is observed before this divergence time. By grouping the ubiquitination proteins into different functional categories, we confirm that many functional constraints like certain molecular functions, protein tissue expression specificity and protein connectivity in protein-protein interaction network enhance the evolutionary conservation of ubiquitination sites. Furthermore, by analyzing the gains of ubiquitination sites at different divergence time and their functional characters, we validate that the emergences of ubiquitination sites at different evolutionary time were also affected by the uncovered functional constraints. The above results suggest that functional constraints on the adaptive evolution of ubiquitination sites increase the opportunity for ubiquitination to synthetically regulate various cellular and developmental processes during evolution.

A Cell-type-resolved Liver Proteome.

Parenchymatous organs consist of multiple cell types, primarily defined as parenchymal cells (PCs) and nonparenchymal cells (NPCs). The cellular characteristics of these organs are not well understood. Proteomic studies facilitate the resolution of the molecular details of different cell types in organs. These studies have significantly extended our knowledge about organogenesis and organ cellular composition. Here, we present an atlas of the cell-type-resolved liver proteome. In-depth proteomics identified 6000 to 8000 gene products (GPs) for each cell type and a total of 10,075 GPs for four cell types. This data set revealed features of the cellular composition of the liver: (1) hepatocytes (PCs) express the least GPs, have a unique but highly homogenous proteome pattern, and execute fundamental liver functions; (2) the division of labor among PCs and NPCs follows a model in which PCs make the main components of pathways, but NPCs trigger the pathways; and (3) crosstalk among NPCs and PCs maintains the PC phenotype. This study presents the liver proteome at cell resolution, serving as a research model for dissecting the cell type constitution and organ features at the molecular level.