Systematically Evaluating Cell-Free DNA Fragmentation Patterns for Cancer Diagnosis and Enhanced Cancer Detection via Integrating Multiple Fragmentation Patterns.

Cell-free DNA (cfDNA) fragmentation patterns have immense potential for early cancer detection. However, the definition of fragmentation varies, ranging from the entire genome to specific genomic regions. These patterns have not been systematically compared, impeding broader research and practical implementation. Here, 1382 plasma cfDNA sequencing samples from 8 cancer types are collected. Considering that cfDNA within open chromatin regions is more susceptible to fragmentation, 10 fragmentation patterns within open chromatin regions as features and employed machine learning techniques to evaluate their performance are examined. All fragmentation patterns demonstrated discernible classification capabilities, with the end motif showing the highest diagnostic value for cross-validation. Combining cross and independent validation results revealed that fragmentation patterns that incorporated both fragment length and coverage information exhibited robust predictive capacities. Despite their diagnostic potential, the predictive power of these fragmentation patterns is unstable. To address this limitation, an ensemble classifier via integrating all fragmentation patterns is developed, which demonstrated notable improvements in cancer detection and tissue-of-origin determination. Further functional bioinformatics investigations on significant feature intervals in the model revealed its impressive ability to identify critical regulatory regions involved in cancer pathogenesis.

Platelet Membrane-Camouflaged Silver Metal-Organic Framework Biomimetic Nanoparticles for the Treatment of Triple-Negative Breast Cancer.

Triple-negative breast cancer (TNBC) is characterized by high malignancy and limited treatment options. Given the pressing need for more effective treatments for TNBC, this study aimed to develop platelet membrane (PM)-camouflaged silver metal-organic framework nanoparticles (PM@MOF-Ag NPs), a biomimetic nanodrug. PM@MOF-Ag NP construction involved the utilization of 2-methylimidazole and silver nitrate to prepare silver metal-organic framework (MOF-Ag) NPs. The PM@MOF-Ag NPs, due to their camouflage, possess excellent blood compatibility, immune escape ability, and a strong affinity for 4T1 tumor cells. This enhances their circulation time in vivo and promotes the aggregation of PM@MOF-Ag NPs at the 4T1 tumor site. Importantly, PM@MOF-Ag NPs demonstrated promising antitumor activity in vitro and in vivo. We further revealed that PM@MOF-Ag NPs induced tumor cell death by overproducing reactive oxygen species and promoting cell apoptosis. Moreover, PM@MOF-Ag NPs enhanced apoptosis by upregulating the ratios of Bax/Bcl-2 and cleaved caspase3/pro-caspase3. Notably, PM@MOF-Ag NPs exhibited no significant organ toxicity, whereas the administration of MOF-Ag NPs resulted in liver inflammation compared to the control group.

PheSeq, a Bayesian deep learning model to enhance and interpret the gene-disease association studies.

Despite the abundance of genotype-phenotype association studies, the resulting association outcomes often lack robustness and interpretations. To address these challenges, we introduce PheSeq, a Bayesian deep learning model that enhances and interprets association studies through the integration and perception of phenotype descriptions. By implementing the PheSeq model in three case studies on Alzheimer's disease, breast cancer, and lung cancer, we identify 1024 priority genes for Alzheimer's disease and 818 and 566 genes for breast cancer and lung cancer, respectively. Benefiting from data fusion, these findings represent moderate positive rates, high recall rates, and interpretation in gene-disease association studies.

Deep centroid: a general deep cascade classifier for biomedical omics data classification.

MOTIVATION: Classification of samples using biomedical omics data is a widely used method in biomedical research. However, these datasets often possess challenging characteristics, including high dimensionality, limited sample sizes, and inherent biases across diverse sources. These factors limit the performance of traditional machine learning models, particularly when applied to independent datasets. RESULTS: To address these challenges, we propose a novel classifier, Deep Centroid, which combines the stability of the nearest centroid classifier and the strong fitting ability of the deep cascade strategy. Deep Centroid is an ensemble learning method with a multi-layer cascade structure, consisting of feature scanning and cascade learning stages that can dynamically adjust the training scale. We apply Deep Centroid to three precision medicine applications-cancer early diagnosis, cancer prognosis, and drug sensitivity prediction-using cell-free DNA fragmentations, gene expression profiles, and DNA methylation data. Experimental results demonstrate that Deep Centroid outperforms six traditional machine learning models in all three applications, showcasing its potential in biological omics data classification. Furthermore, functional annotations reveal that the features scanned by the model exhibit biological significance, indicating its interpretability from a biological perspective. Our findings underscore the promising application of Deep Centroid in the classification of biomedical omics data, particularly in the field of precision medicine. AVAILABILITY AND IMPLEMENTATION: Deep Centroid is available at both github (github.com/xiexiexiekuan/DeepCentroid) and Figshare (https://figshare.com/articles/software/Deep_Centroid_A_General_Deep_Cascade_Classifier_for_Biomedical_Omics_Data_Classification/24993516).

Bioengineered stem cell membrane functionalized nanoparticles combine anti-inflammatory and antimicrobial properties for sepsis treatment.

BACKGROUND: Sepsis is a syndrome of physiological, pathological and biochemical abnormalities caused by infection. Although the mortality rate is lower than before, many survivors have persistent infection, which means sepsis calls for new treatment. After infection, inflammatory mediators were largely released into the blood, leading to multiple organ dysfunction. Therefore, anti-infection and anti-inflammation are critical issues in sepsis management. RESULTS: Here, we successfully constructed a novel nanometer drug loading system for sepsis management, FZ/MER-AgMOF@Bm. The nanoparticles were modified with LPS-treated bone marrow mesenchymal stem cell (BMSC) membrane, and silver metal organic framework (AgMOF) was used as the nanocore for loading FPS-ZM1 and meropenem which was delivery to the infectious microenvironments (IMEs) to exert dual anti-inflammatory and antibacterial effects. FZ/MER-AgMOF@Bm effectively alleviated excessive inflammatory response and eliminated bacteria. FZ/MER-AgMOF@Bm also played an anti-inflammatory role by promoting the polarization of macrophages to M2. When sepsis induced by cecal ligation and puncture (CLP) challenged mice was treated, FZ/MER-AgMOF@Bm could not only reduce the levels of pro-inflammatory factors and lung injury, but also help to improve hypothermia caused by septic shock and prolong survival time. CONCLUSIONS: Together, the nanoparticles played a role in combined anti-inflammatory and antimicrobial properties, alleviating cytokine storm and protecting vital organ functions, could be a potential new strategy for sepsis management.

CRAG: de novo characterization of cell-free DNA fragmentation hotspots in plasma whole-genome sequencing.

The fine-scale cell-free DNA fragmentation patterns in early-stage cancers are poorly understood. We developed a de novo approach to characterize the cell-free DNA fragmentation hotspots from plasma whole-genome sequencing. Hotspots are enriched in open chromatin regions, and, interestingly, 3'end of transposons. Hotspots showed global hypo-fragmentation in early-stage liver cancers and are associated with genes involved in the initiation of hepatocellular carcinoma and associated with cancer stem cells. The hotspots varied across multiple early-stage cancers and demonstrated high performance for the diagnosis and identification of tissue-of-origin in early-stage cancers. We further validated the performance with a small number of independent case-control-matched early-stage cancer samples.

Discovering common pathogenetic processes between COVID-19 and sepsis by bioinformatics and system biology approach.

Corona Virus Disease 2019 (COVID-19), an acute respiratory infectious disease caused by severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2), has spread rapidly worldwide, resulting in a pandemic with a high mortality rate. In clinical practice, we have noted that many critically ill or critically ill patients with COVID-19 present with typical sepsis-related clinical manifestations, including multiple organ dysfunction syndrome, coagulopathy, and septic shock. In addition, it has been demonstrated that severe COVID-19 has some pathological similarities with sepsis, such as cytokine storm, hypercoagulable state after blood balance is disrupted and neutrophil dysfunction. Considering the parallels between COVID-19 and non-SARS-CoV-2 induced sepsis (hereafter referred to as sepsis), the aim of this study was to analyze the underlying molecular mechanisms between these two diseases by bioinformatics and a systems biology approach, providing new insights into the pathogenesis of COVID-19 and the development of new treatments. Specifically, the gene expression profiles of COVID-19 and sepsis patients were obtained from the Gene Expression Omnibus (GEO) database and compared to extract common differentially expressed genes (DEGs). Subsequently, common DEGs were used to investigate the genetic links between COVID-19 and sepsis. Based on enrichment analysis of common DEGs, many pathways closely related to inflammatory response were observed, such as Cytokine-cytokine receptor interaction pathway and NF-kappa B signaling pathway. In addition, protein-protein interaction networks and gene regulatory networks of common DEGs were constructed, and the analysis results showed that ITGAM may be a potential key biomarker base on regulatory analysis. Furthermore, a disease diagnostic model and risk prediction nomogram for COVID-19 were constructed using machine learning methods. Finally, potential therapeutic agents, including progesterone and emetine, were screened through drug-protein interaction networks and molecular docking simulations. We hope to provide new strategies for future research and treatment related to COVID-19 by elucidating the pathogenesis and genetic mechanisms between COVID-19 and sepsis.

Multi-way relation-enhanced hypergraph representation learning for anti-cancer drug synergy prediction.

MOTIVATION: Drug combinations have exhibited promise in treating cancers with less toxicity and fewer adverse reactions. However, in vitro screening of synergistic drug combinations is time-consuming and labor-intensive because of the combinatorial explosion. Although a number of computational methods have been developed for predicting synergistic drug combinations, the multi-way relations between drug combinations and cell lines existing in drug synergy data have not been well exploited. RESULTS: We propose a multi-way relation-enhanced hypergraph representation learning method to predict anti-cancer drug synergy, named HypergraphSynergy. HypergraphSynergy formulates synergistic drug combinations over cancer cell lines as a hypergraph, in which drugs and cell lines are represented by nodes and synergistic drug-drug-cell line triplets are represented by hyperedges, and leverages the biochemical features of drugs and cell lines as node attributes. Then, a hypergraph neural network is designed to learn the embeddings of drugs and cell lines from the hypergraph and predict drug synergy. Moreover, the auxiliary task of reconstructing the similarity networks of drugs and cell lines is considered to enhance the generalization ability of the model. In the computational experiments, HypergraphSynergy outperforms other state-of-the-art synergy prediction methods on two benchmark datasets for both classification and regression tasks and is applicable to unseen drug combinations or cell lines. The studies revealed that the hypergraph formulation allows us to capture and explain complex multi-way relations of drug combinations and cell lines, and also provides a flexible framework to make the best use of diverse information. AVAILABILITY AND IMPLEMENTATION: The source data and codes of HypergraphSynergy can be freely downloaded from https://github.com/liuxuan666/HypergraphSynergy. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Advancing Prediction of Risk of Intraoperative Massive Blood Transfusion in Liver Transplantation With Machine Learning Models. A Multicenter Retrospective Study.

Background: Liver transplantation surgery is often accompanied by massive blood loss and massive transfusion (MT), while MT can cause many serious complications related to high mortality. Therefore, there is an urgent need for a model that can predict the demand for MT to reduce the waste of blood resources and improve the prognosis of patients. Objective: To develop a model for predicting intraoperative massive blood transfusion in liver transplantation surgery based on machine learning algorithms. Methods: A total of 1,239 patients who underwent liver transplantation surgery in three large grade lll-A general hospitals of China from March 2014 to November 2021 were included and analyzed. A total of 1193 cases were randomly divided into the training set (70%) and test set (30%), and 46 cases were prospectively collected as a validation set. The outcome of this study was an intraoperative massive blood transfusion. A total of 27 candidate risk factors were collected, and recursive feature elimination (RFE) was used to select key features based on the Categorical Boosting (CatBoost) model. A total of ten machine learning models were built, among which the three best performing models and the traditional logistic regression (LR) method were prospectively verified in the validation set. The Area Under the Receiver Operating Characteristic Curve (AUROC) was used for model performance evaluation. The Shapley additive explanation value was applied to explain the complex ensemble learning models. Results: Fifteen key variables were screened out, including age, weight, hemoglobin, platelets, white blood cells count, activated partial thromboplastin time, prothrombin time, thrombin time, direct bilirubin, aspartate aminotransferase, total protein, albumin, globulin, creatinine, urea. Among all algorithms, the predictive performance of the CatBoost model (AUROC: 0.810) was the best. In the prospective validation cohort, LR performed far less well than other algorithms. Conclusion: A prediction model for massive blood transfusion in liver transplantation surgery was successfully established based on the CatBoost algorithm, and a certain degree of generalization verification is carried out in the validation set. The model may be superior to the traditional LR model and other algorithms, and it can more accurately predict the risk of massive blood transfusions and guide clinical decision-making.

Combined Microbiome and Metabolome Analysis Reveals a Novel Interplay Between Intestinal Flora and Serum Metabolites in Lung Cancer.

As the leading cause of cancer death, lung cancer seriously endangers human health and quality of life. Although many studies have reported the intestinal microbial composition of lung cancer, little is known about the interplay between intestinal microbiome and metabolites and how they affect the development of lung cancer. Herein, we combined 16S ribosomal RNA (rRNA) gene sequencing and liquid chromatography-mass spectrometry (LC-MS) technology to analyze intestinal microbiota composition and serum metabolism profile in a cohort of 30 lung cancer patients with different stages and 15 healthy individuals. Compared with healthy people, we found that the structure of intestinal microbiota in lung cancer patients had changed significantly (Adonis, p = 0.021). In order to determine how intestinal flora affects the occurrence and development of lung cancer, the Spearman rank correlation test was used to find the connection between differential microorganisms and differential metabolites. It was found that as thez disease progressed, L-valine decreased. Correspondingly, the abundance of Lachnospiraceae_UCG-006, the genus with the strongest association with L-valine, also decreased in lung cancer groups. Correlation analysis showed that the gut microbiome and serum metabolic profile had a strong synergy, and Lachnospiraceae_UCG-006 was closely related to L-valine. In summary, this study described the characteristics of intestinal flora and serum metabolic profiles of lung cancer patients with different stages. It revealed that lung cancer may be the result of the mutual regulation of L-valine and Lachnospiraceae_UCG-006 through the aminoacyl-tRNA biosynthesis pathway, and proposed that L-valine may be a potential marker for the diagnosis of lung cancer.

Systems Pharmacology-Based Precision Therapy and Drug Combination Discovery for Breast Cancer.

Breast cancer (BC) is a common disease and one of the main causes of death in females worldwide. In the omics era, researchers have used various high-throughput sequencing technologies to accumulate massive amounts of biomedical data and reveal an increasing number of disease-related mutations/genes. It is a major challenge to use these data effectively to find drugs that may protect human health. In this study, we combined the GeneRank algorithm and gene dependency network to propose a precision drug discovery strategy that can recommend drugs for individuals and screen existing drugs that could be used to treat different BC subtypes. We used this strategy to screen four BC subtype-specific drug combinations and verified the potential activity of combining gefitinib and irinotecan in triple-negative breast cancer (TNBC) through in vivo and in vitro experiments. The results of cell and animal experiments demonstrated that the combination of gefitinib and irinotecan can significantly inhibit the growth of TNBC tumour cells. The results also demonstrated that this systems pharmacology-based precision drug discovery strategy effectively identified important disease-related genes in individuals and special groups, which supports its efficiency, high reliability, and practical application value in drug discovery.

MNBDR: A Module Network Based Method for Drug Repositioning.

Drug repurposing/repositioning, which aims to find novel indications for existing drugs, contributes to reducing the time and cost for drug development. For the recent decade, gene expression profiles of drug stimulating samples have been successfully used in drug repurposing. However, most of the existing methods neglect the gene modules and the interactions among the modules, although the cross-talks among pathways are common in drug response. It is essential to develop a method that utilizes the cross-talks information to predict the reliable candidate associations. In this study, we developed MNBDR (Module Network Based Drug Repositioning), a novel method that based on module network to screen drugs. It integrated protein-protein interactions and gene expression profile of human, to predict drug candidates for diseases. Specifically, the MNBDR mined dense modules through protein-protein interaction (PPI) network and constructed a module network to reveal cross-talks among modules. Then, together with the module network, based on existing gene expression data set of drug stimulation samples and disease samples, we used random walk algorithms to capture essential modules in disease development and proposed a new indicator to screen potential drugs for a given disease. Results showed MNBDR could provide better performance than popular methods. Moreover, functional analysis of the essential modules in the network indicated our method could reveal biological mechanism in drug response.

DNA Methylation Biomarkers Predict Objective Responses to PD-1/PD-L1 Inhibition Blockade.

Immune checkpoint inhibitor (ICI) treatment could bring long-lasting clinical benefits to patients with metastatic cancer. However, only a small proportion of patients respond to PD-1/PD-L1 blockade, so predictive biomarkers are needed. Here, based on DNA methylation profiles and the objective response rates (ORRs) of PD-1/PD-L1 inhibition therapy, we identified 269 CpG sites and developed an initial CpG-based model by Lasso to predict ORRs. Notably, as measured by the area under the receiver operating characteristic curve (AUC), our model can produce better performance (AUC = 0.92) than both a model based on tumor mutational burden (TMB) (AUC = 0.77) and a previously reported TMB model (AUC = 0.71). In addition, most CpGs also have additional synergies with TMB, which can achieve a higher prediction accuracy when joined with TMB. Furthermore, we identified CpGs that are associated with TMB at the individual level. DNA methylation modules defined by protein networks, Kyoto Encylopedia of Genes and Genomes (KEGG) pathways, and ligand-receptor gene pairs are also associated with ORRs. This method suggested novel immuno-oncology targets that might be beneficial when combined with PD-1/PD-L1 blockade. Thus, DNA methylation studies might hold great potential for individualized PD1/PD-L1 blockade or combinatory therapy.

DNA Methylation Module Network-Based Prognosis and Molecular Typing of Cancer.

Achieving cancer prognosis and molecular typing is critical for cancer treatment. Previous studies have identified some gene signatures for the prognosis and typing of cancer based on gene expression data. Some studies have shown that DNA methylation is associated with cancer development, progression, and metastasis. In addition, DNA methylation data are more stable than gene expression data in cancer prognosis. Therefore, in this work, we focused on DNA methylation data. Some prior researches have shown that gene modules are more reliable in cancer prognosis than are gene signatures and that gene modules are not isolated. However, few studies have considered cross-talk among the gene modules, which may allow some important gene modules for cancer to be overlooked. Therefore, we constructed a gene co-methylation network based on the DNA methylation data of cancer patients, and detected the gene modules in the co-methylation network. Then, by permutation testing, cross-talk between every two modules was identified; thus, the module network was generated. Next, the core gene modules in the module network of cancer were identified using the K-shell method, and these core gene modules were used as features to study the prognosis and molecular typing of cancer. Our method was applied in three types of cancer (breast invasive carcinoma, skin cutaneous melanoma, and uterine corpus endometrial carcinoma). Based on the core gene modules identified by the constructed DNA methylation module networks, we can distinguish not only the prognosis of cancer patients but also use them for molecular typing of cancer. These results indicated that our method has important application value for the diagnosis of cancer and may reveal potential carcinogenic mechanisms.

Prediction of Bone Metastasis in Breast Cancer Based on Minimal Driver Gene Set in Gene Dependency Network.

Bone is the most frequent organ for breast cancer metastasis, and thus it is essential to predict the bone metastasis of breast cancer. In our work, we constructed a gene dependency network based on the hypothesis that the relation between one gene and the risk of bone metastasis might be affected by another gene. Then, based on the structure controllability theory, we mined the driver gene set which can control the whole network in the gene dependency network, and the signature genes were selected from them. Survival analysis showed that the signature could distinguish the bone metastasis risks of cancer patients in the test data set and independent data set. Besides, we used the signature genes to construct a centroid classifier. The results showed that our method is effective and performed better than published methods.

An Ensemble Strategy to Predict Prognosis in Ovarian Cancer Based on Gene Modules.

Due to the high heterogeneity and complexity of cancer, it is still a challenge to predict the prognosis of cancer patients. In this work, we used a clustering algorithm to divide patients into different subtypes in order to reduce the heterogeneity of the cancer patients in each subtype. Based on the hypothesis that the gene co-expression network may reveal relationships among genes, some communities in the network could influence the prognosis of cancer patients and all the prognosis-related communities could fully reveal the prognosis of cancer patients. To predict the prognosis for cancer patients in each subtype, we adopted an ensemble classifier based on the gene co-expression network of the corresponding subtype. Using the gene expression data of ovarian cancer patients in TCGA (The Cancer Genome Atlas), three subtypes were identified. Survival analysis showed that patients in different subtypes had different survival risks. Three ensemble classifiers were constructed for each subtype. Leave-one-out and independent validation showed that our method outperformed control and literature methods. Furthermore, the function annotation of the communities in each subtype showed that some communities were cancer-related. Finally, we found that the current drug targets can partially support our method.

High CD3D/CD4 ratio predicts better survival in muscle-invasive bladder cancer.

Background: Bladder cancer is a common malignancy that affects the human urinary tract. Muscle-invasive bladder cancer (MIBC) is aggressive and has poor prognosis. Previous studies have reported that the tumor-infiltrating lymphocytes (TILs) were associated with MIBC outcome; however, inconsistency remains and mRNA level TIL markers' prognostic significance in MIBC is unclear.  Materials and methods: In the present study, we reanalyzed data from four public datasets (the Cancer Genome Atlas for investigation; and CIT, GSE5287, and GSE31684 for validation) to examine the prognostic significance of CD3D, CD4, CD8A, CD3D/CD4 and CD3D/CD8A in MIBC.  Results: We found that the CD3D/CD4 ratio was a stable independent prognostic factor in MIBC (beta = -0.87, P = 0.025); high CD3D/CD4 ratio predicted better survival in MIBC, and the power of this association was much stronger in basal-squamous tumors (beta = -4.73, P = 2.67E-06). We also noted that the CD4 expression was significantly higher than CD3D (P < 0.05), indicating the presence of CD3-CD4+ cells which could be immune-suppressing. Conclusion: The CD3D/CD4 ratio can be viewed as a prognostic marker and a rough measurement for the interaction between immune-effecting CD3+ TILs and immune-suppressing CD3-CD4+ cells in MIBC, and this interaction may play a particularly important role in anti-cancer immunity in basal-squamous tumors as it has a very strong association with survival in this subtype, and may be used to select potential responders to immunotherapy.

Identification of Cancer Hallmarks Based on the Gene Co-expression Networks of Seven Cancers.

Identifying the hallmarks of cancer is essential for cancer research, and the genes involved in cancer hallmarks are likely to be cancer drivers. However, there is no appropriate method in the current literature for identifying genetic cancer hallmarks, especially considering the interrelationships among the genes. Here, we hypothesized that "dense clusters" (or "communities") in the gene co-expression networks of cancer patients may represent functional units regarding cancer formation and progression, and the communities present in the co-expression networks of multiple types of cancer may be cancer hallmarks. Consequently, we mined the conserved communities in the gene co-expression networks of seven cancers in order to identify candidate hallmarks. Functional annotation of the communities showed that they were mainly related to immune response, the cell cycle and the biological processes that maintain basic cellular functions. Survival analysis using the genes involved in the conserved communities verified that two of these hallmarks could predict the survival risks of cancer patients in multiple types of cancer. Furthermore, the genes involved in these hallmarks, one of which was related to the cell cycle, could be useful in screening for cancer drugs.

PheWAS-Based Systems Genetics Methods for Anti-Breast Cancer Drug Discovery.

Breast cancer is a high-risk disease worldwide. For such complex diseases that are induced by multiple pathogenic genes, determining how to establish an effective drug discovery strategy is a challenge. In recent years, a large amount of genetic data has accumulated, particularly in the genome-wide identification of disorder genes. However, understanding how to use these data efficiently for pathogenesis elucidation and drug discovery is still a problem because the gene⁻disease links that are identified by high-throughput techniques such as phenome-wide association studies (PheWASs) are usually too weak to have biological significance. Systems genetics is a thriving area of study that aims to understand genetic interactions on a genome-wide scale. In this study, we aimed to establish two effective strategies for identifying breast cancer genes based on the systems genetics algorithm. As a result, we found that the GeneRank-based strategy, which combines the prognostic phenotype-based gene-dependent network with the phenotypic-related PheWAS data, can promote the identification of breast cancer genes and the discovery of anti-breast cancer drugs.

Identifying cancer prognostic modules by module network analysis.

BACKGROUND: The identification of prognostic genes that can distinguish the prognostic risks of cancer patients remains a significant challenge. Previous works have proven that functional gene sets were more reliable for this task than the gene signature. However, few works have considered the cross-talk among functional gene sets, which may result in neglecting important prognostic gene sets for cancer. RESULTS: Here, we proposed a new method that considers both the interactions among modules and the prognostic correlation of the modules to identify prognostic modules in cancers. First, dense sub-networks in the gene co-expression network of cancer patients were detected. Second, cross-talk between every two modules was identified by a permutation test, thus generating the module network. Third, the prognostic correlation of each module was evaluated by the resampling method. Then, the GeneRank algorithm, which takes the module network and the prognostic correlations of all the modules as input, was applied to prioritize the prognostic modules. Finally, the selected modules were validated by survival analysis in various data sets. Our method was applied in three kinds of cancers, and the results show that our method succeeded in identifying prognostic modules in all the three cancers. In addition, our method outperformed state-of-the-art methods. Furthermore, the selected modules were significantly enriched with known cancer-related genes and drug targets of cancer, which may indicate that the genes involved in the modules may be drug targets for therapy. CONCLUSIONS: We proposed a useful method to identify key modules in cancer prognosis and our prognostic genes may be good candidates for drug targets.