Identification of a 6-gene signature predicting prognosis for colorectal cancer

Identification and Validation of Novel Metastasis-Related Immune Gene Signature in Breast Cancer

Background

Distant metastasis remains the leading cause of death among patients with breast cancer (BRCA). The process of cancer metastasis involves multiple mechanisms, including compromised immune system. However, not all genes involved in immune function have been comprehensively identified.

Methods

Firstly 1623 BRCA samples, including transcriptome sequencing and clinical information, were acquired from Gene Expression Omnibus (GSE102818, GSE45255, GSE86166) and The Cancer Genome Atlas-BRCA (TCGA-BRCA) dataset. Subsequently, weighted gene co-expression network analysis (WGCNA) was performed using the GSE102818 dataset to identify the most relevant module to the metastasis of BRCA. Besides, ConsensusClusterPlus was applied to divide TCGA-BRCA patients into two subgroups (G1 and G2). In the meantime, the least absolute shrinkage and selection operator (LASSO) regression analysis was used to construct a metastasis-related immune genes (MRIGs)_score to predict the metastasis and progression of cancer. Importantly, the expression of vital genes was validated through reverse transcription quantitative polymerase chain reaction (RT-qPCR) and immunohistochemistry (IHC).

Results

The expression pattern of 76 MRIGs screened by WGCNA divided TCGA-BRCA patients into two subgroups (G1 and G2), and the prognosis of G1 group was worse. Also, G1 exhibited a higher mRNA expression level based on stemness index score and Tumor Immune Dysfunction and Exclusion score. In addition, higher MRIGs_score represented the higher probability of progression in BRCA patients. It was worth mentioning that the patients in the G1 group had a high MRIGs_score than those in the G2 group. Importantly, the results of RT-qPCR and IHC demonstrated that fasciculation and elongation protein zeta 1 (FEZ1) and insulin-like growth factor 2 receptor (IGF2R) were risk factors, while interleukin (IL)-1 receptor antagonist (IL1RN) was a protective factor.

Conclusion

Our study revealed a prognostic model composed of eight immune related genes that could predict the metastasis and progression of BRCA. Higher score represented higher metastasis probability. Besides, the consistency of key genes in BRCA tissue and bioinformatics analysis results from mRNA and protein levels was verified.

Potential role of the Eph/ephrin system in colorectal cancer: emerging druggable molecular targets

The Eph/ephrin system regulates many developmental processes and adult tissue homeostasis. In colorectal cancer (CRC), it is involved in different processes including tumorigenesis, tumor angiogenesis, metastasis development, and cancer stem cell regeneration. However, conflicting data regarding Eph receptors in CRC, especially in its putative role as an oncogene or a suppressor gene, make the precise role of Eph-ephrin interaction confusing in CRC development. In this review, we provide an overview of the literature and highlight evidence that collaborates with these ambiguous roles of the Eph/ephrin system in CRC, as well as the molecular findings that represent promising therapeutic targets.

Dissecting multifunctional roles of forkhead box transcription factor D1 in cancers

As a member of the forkhead box (FOX) family of transcription factors (TF), FOXD1 has recently been implicated as a crucial regulator in a variety of human cancers. Accumulating evidence has established dysregulated and aberrant FOXD1 signaling as a prominent feature in cancer development and progression. However, there is a lack of systematic review on this topic. Here, we summarized the present understanding of FOXD1 functions in cancer biology and reviewed the downstream targets and upstream regulatory mechanisms of FOXD1 as well as the related signaling pathways within the context of current reports. We highlighted the functional features of FOXD1 in cancers to identify the future research consideration of this multifunctional transcription factor and potential therapeutic strategies targeting its oncogenic activity.

A cross-cohort computational framework to trace tumor tissue-of-origin based on RNA sequencing

Carcinoma of unknown primary (CUP) is a type of metastatic cancer with tissue-of-origin (TOO) unidentifiable by traditional methods. CUP patients typically have poor prognosis but therapy targeting the original cancer tissue can significantly improve patients' prognosis. Thus, it's critical to develop accurate computational methods to infer cancer TOO. While qPCR or microarray-based methods are effective in inferring TOO for most cancer types, the overall prediction accuracy is yet to be improved. In this study, we propose a cross-cohort computational framework to trace TOO of 32 cancer types based on RNA sequencing (RNA-seq). Specifically, we employed logistic regression models to select 80 genes for each cancer type to create a combined 1356-gene set, based on transcriptomic data from 9911 tissue samples covering the 32 cancer types with known TOO from the Cancer Genome Atlas (TCGA). The selected genes are enriched in both tissue-specific and tissue-general functions. The cross-validation accuracy of our framework reaches 97.50% across all cancer types. Furthermore, we tested the performance of our model on the TCGA metastatic dataset and International Cancer Genome Consortium (ICGC) dataset, achieving an accuracy of 91.09% and 82.67%, respectively, despite the differences in experiment procedures and pipelines. In conclusion, we developed an accurate yet robust computational framework for identifying TOO, which holds promise for clinical applications. Our code is available at http://github.com/wangbo00129/classifybysklearn .

Identification of Novel Biomarkers for Response to Preoperative Chemoradiation in Locally Advanced Rectal Cancer with Genetic Algorithm-Based Gene Selection

BACKGROUND

The conventional treatment for patients with locally advanced colorectal tumors is preoperative chemo-radiotherapy (PCRT) preceding surgery. This treatment strategy has some long-term side effects, and some patients do not respond to it. Therefore, an evaluation of biomarkers that may help predict patients' response to PCRT is essential.

METHODS

We took advantage of genetic algorithm to search the space of possible combinations of features to choose subsets of genes that would yield convenient performance in differentiating PCRT responders from non-responders using a logistic regression model as our classifier.

RESULTS

We developed two gene signatures; first, to achieve the maximum prediction accuracy, the algorithm yielded 39 genes, and then, aiming to reduce the feature numbers as much as possible (while maintaining acceptable performance), a 5-gene signature was chosen. The performance of the two gene signatures was (accuracy = 0.97 and 0.81, sensitivity = 0.96 and 0.83, and specificity = 86 and 0.77) using a logistic regression classifier. Through analyzing bias and variance decomposition of the model error, we further investigated the involved genes by discovering and validating another 28-gene signature which possibly points towards two different sub-systems involved in the response of the patients to treatment.

CONCLUSIONS

Using genetic algorithm as our gene selection method, we have identified two groups of genes that can differentiate PCRT responders from non-responders in patients of the studied dataset with considerable performance.

IMPACT

After passing standard requirements, our gene signatures may be applicable as a robust and effective PCRT response prediction tool for colorectal cancer patients in clinical settings and may also help future studies aiming to further investigate involved pathways gain a clearer picture for the course of their research.

From Omic Layers to Personalized Medicine in Colorectal Cancer: The Road Ahead

Colorectal cancer is a major health concern since it is a highly diagnosed cancer and the second cause of death among cancers. Thus, the most suitable biomarkers for its diagnosis, prognosis, and treatment have been studied to improve and personalize the prevention and clinical management of colorectal cancer. The emergence of omic techniques has provided a great opportunity to better study CRC and make personalized medicine feasible. In this review, we will try to summarize how the analysis of the omic layers can be useful for personalized medicine and the existing difficulties. We will discuss how single and multiple omic layer analyses have been used to improve the prediction of the risk of CRC and its outcomes and how to overcome the challenges in the use of omic layers in personalized medicine.

Functional Proteomic Profiling Analysis in Four Major Types of Gastrointestinal Cancers

Gastrointestinal (GI) cancer accounts for one in four cancer cases and one in three cancer-related deaths globally. A deeper understanding of cancer development mechanisms can be applied to cancer medicine. Comprehensive sequencing applications have revealed the genomic landscapes of the common types of human cancer, and proteomics technology has identified protein targets and signalling pathways related to cancer growth and progression. This study aimed to explore the functional proteomic profiles of four major types of GI tract cancer based on The Cancer Proteome Atlas (TCPA). We provided an overview of functional proteomic heterogeneity by performing several approaches, including principal component analysis (PCA), partial least squares discriminant analysis (PLS-DA), t-stochastic neighbour embedding (t-SNE) analysis, and hierarchical clustering analysis in oesophageal carcinoma (ESCA), stomach adenocarcinoma (STAD), colon adenocarcinoma (COAD), and rectum adenocarcinoma (READ) tumours, to gain a system-wide understanding of the four types of GI cancer. The feature selection approach, mutual information feature selection (MIFS) method, was conducted to screen candidate protein signature subsets to better distinguish different cancer types. The potential clinical implications of candidate proteins in terms of tumour progression and prognosis were also evaluated based on TCPA and The Cancer Genome Atlas (TCGA) databases. The results suggested that functional proteomic profiling can identify different patterns among the four types of GI cancers and provide candidate proteins for clinical diagnosis and prognosis evaluation. We also highlighted the application of feature selection approaches in high-dimensional biological data analysis. Overall, this study could improve the understanding of the complexity of cancer phenotypes and genotypes and thus be applied to cancer medicine.

Prognostic model development for classification of colorectal adenocarcinoma by using machine learning model based on feature selection technique boruta

Colorectal cancer (CRC) is the third most prevalent cancer type and accounts for nearly one million deaths worldwide. The CRC mRNA gene expression datasets from TCGA and GEO (GSE144259, GSE50760, and GSE87096) were analyzed to find the significant differentially expressed genes (DEGs). These significant genes were further processed for feature selection through boruta and the confirmed features of importance (genes) were subsequently used for ML-based prognostic classification model development. These genes were analyzed for survival and correlation analysis between final genes and infiltrated immunocytes. A total of 770 CRC samples were included having 78 normal and 692 tumor tissue samples. 170 significant DEGs were identified after DESeq2 analysis along with the topconfects R package. The 33 confirmed features of importance-based RF prognostic classification model have given accuracy, precision, recall, and f1-score of 100% with 0% standard deviation. The overall survival analysis had finalized GLP2R and VSTM2A genes that were significantly downregulated in tumor samples and had a strong correlation with immunocyte infiltration. The involvement of these genes in CRC prognosis was further confirmed on the basis of their biological function and literature analysis. The current findings indicate that GLP2R and VSTM2A may play a significant role in CRC progression and immune response suppression.

Colon cancer transcriptome

Over the last four decades, methodological innovations have continuously changed transcriptome profiling. It is now feasible to sequence and quantify the transcriptional outputs of individual cells or thousands of samples using RNA sequencing (RNA-seq). These transcriptomes serve as a connection between cellular behaviors and their underlying molecular mechanisms, such as mutations. This relationship, in the context of cancer, provides a chance to unravel tumor complexity and heterogeneity and uncover novel biomarkers or treatment options. Since colon cancer is one of the most frequent malignancies, its prognosis and diagnosis seem to be critical. The transcriptome technology is developing for an earlier and more accurate diagnosis of cancer which can provide better protectivity and prognostic utility to medical teams and patients. A transcriptome is a whole set of expressed coding and non-coding RNAs in an individual or cell population. The cancer transcriptome includes RNA-based changes. The combined genome and transcriptome of a patient may provide a comprehensive picture of their cancer, and this information is beginning to affect treatment decision-making in real-time. A full assessment of the transcriptome of colon (colorectal) cancer has been assessed in this review paper based on risk factors such as age, obesity, gender, alcohol use, race, and also different stages of cancer, as well as non-coding RNAs like circRNAs, miRNAs, lncRNAs, and siRNAs. Similarly, they have been examined independently in the transcriptome study of colon cancer.

Iroquois Family Genes in Gastric Carcinogenesis: A Comprehensive Review

Gastric cancer (GC) is the fifth leading cause of cancer-associated death worldwide, accounting for 768,793 related deaths and 1,089,103 new cases in 2020. Despite diagnostic advances, GC is often detected in late stages. Through a systematic literature search, this study focuses on the associations between the Iroquois gene family and GC. Accumulating evidence indicates that Iroquois genes are involved in the regulation of various physiological and pathological processes, including cancer. To date, information about Iroquois genes in GC is very limited. In recent years, the expression and function of Iroquois genes examined in different models have suggested that they play important roles in cell and cancer biology, since they were identified to be related to important signaling pathways, such as wingless, hedgehog, mitogen-activated proteins, fibroblast growth factor, TGFβ, and the PI3K/Akt and NF-kB pathways. In cancer, depending on the tumor, Iroquois genes can act as oncogenes or tumor suppressor genes. However, in GC, they seem to mostly act as tumor suppressor genes and can be regulated by several mechanisms, including methylation, microRNAs and important GC-related pathogens. In this review, we provide an up-to-date review of the current knowledge regarding Iroquois family genes in GC.

Regularized survival learning and cross-database analysis enabled identification of colorectal cancer prognosis-related immune genes

Colon adenocarcinoma is the most common type of colorectal cancer. The prognosis of advanced colorectal cancer patients who received treatment is still very poor. Therefore, identifying new biomarkers for prognosis prediction has important significance for improving treatment strategies. However, the power of biomarker analyses was limited by the used sample size of individual database. In this study, we combined Genotype-Tissue Expression (GTEx) and The Cancer Genome Atlas (TCGA) databases to expand the number of healthy tissue samples. We screened differentially expressed genes between the GTEx healthy samples and TCGA tumor samples. Subsequently, we applied least absolute shrinkage and selection operator (LASSO) regression and multivariate Cox analysis to identify nine prognosis-related immune genes: ANGPTL4, IDO1, NOX1, CXCL3, LTB4R, IL1RL2, CD72, NOS2, and NUDT6. We computed the risk scores of samples based on the expression levels of these genes and divided patients into high- and low-risk groups according to this risk score. Survival analysis results showed a significant difference in survival rate between the two risk groups. The high-risk group had a significantly lower overall survival rate and poorer prognosis. We found the receiver operating characteristic based on the risk score was showed to accurately predict patients' prognosis. These prognosis-related immune genes may be potential biomarkers for colorectal cancer diagnosis and treatment. Our open-source code is freely available from GitHub at https://github.com/gutmicrobes/Prognosis-model.git.

Impact of buffer composition on biochemical, morphological and mechanical parameters: A tare before dielectrophoretic cell separation and isolation

Dielectrophoresis (DEP) represents an electrokinetic approach for discriminating and separating suspended cells based on their intrinsic dielectric characteristics without the need for labeling procedure. A good practice, beyond the physical and engineering components, is the selection of a buffer that does not hinder cellular and biochemical parameters as well as cell recovery. In the present work the impact of four buffers on biochemical, morphological, and mechanical parameters was evaluated in two different cancer cell lines (Caco-2 and K562). Specifically, MTT ([3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide]) assay along with flow cytometry analysis were used to evaluate the occurring changes in terms of cell viability, morphology, and granulocyte stress formation, all factors directly influencing DEP sorting capability. Quantitative real-time PCR (qRT-PCR) was instead employed to evaluate the gene expression levels of interleukin-6 (IL-6) and inducible nitric oxide synthase (iNOS), two well-known markers of inflammation and oxidative stress, respectively. An additional marker representing an index of cellular metabolic status, i.e. the expression of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) gene, was also evaluated. Among the four buffers considered, two resulted satisfactory in terms of cell viability and growth recovery (24 h), with no significant changes in cell morphology for up to 1 h in suspension. Of note, gene expression analysis showed that in both cell lines the apparently non-cytotoxic buffers significantly modulated IL-6, iNOS, and GAPDH markers, underlining the importance to deeply investigate the molecular and biochemical changes occurring during the analysis, even at apparently non-toxic conditions. The selection of a useful buffer for the separation and analysis of cells without labeling procedures, preserving cell status, represents a key factor for DEP analysis, giving the opportunity to further use cells for additional analysis.

A novel signature model based on mitochondrial-related genes for predicting survival of colon adenocarcinoma

Background

Colon cancer is the foremost reason of cancer-related mortality worldwide. Colon adenocarcinoma constitutes 90% of colon cancer, and most patients with colon adenocarcinoma (COAD) are identified until advanced stage. With the emergence of an increasing number of novel pathogenic mechanisms and treatments, the role of mitochondria in the development of cancer, has been studied and reported with increasing frequency.

Methods

We systematically analyzed the effect of mitochondria-related genes in COAD utilizing RNA sequencing dataset from The Cancer Genome Atlas database and 1613 mitochondrial function-related genes from MitoMiner database. Our approach consisted of differentially expressed gene, gene set enrichment analysis, gene ontology terminology, Kyoto Encyclopedia of Genes and Genomes, independent prognostic analysis, univariate and multivariate analysis, Kaplan–Meier survival analysis, immune microenvironment correlation analysis, and Cox regression analysis.

Results

Consequently, 8 genes were identified to construct 8 mitochondrial-related gene model by applying Cox regression analysis, CDC25C, KCNJ11, NOL3, P4HA1, QSOX2, Trap1, DNAJC28, and ATCAY. Meanwhile, we assessed the connection between this model and clinical parameters or immune microenvironment. Risk score was an independent predictor for COAD patients’ survival with an AUC of 0.687, 0.752 and 0.762 at 1-, 3- and 5-year in nomogram, respectively. The group with the highest risk score had the lowest survival rate and the worst clinical stages. Additionally, its predictive capacity was validated in GSE39582 cohort.

Conclusion

In summary, we established a prognostic pattern of mitochondrial-related genes, which can predict overall survival in COAD, which may enable a more optimized approach for the clinical treatment and scientific study of COAD. This gene signature model has the potential to improve prognosis and treatment for COAD patients in the future, and to be widely implemented in clinical settings. The utilization of this mitochondrial-related gene signature model may be benefit in the treatments and medical decision-making of COAD.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12911-022-02020-3.

Identification and validation of an inflammation-related lncRNAs signature for improving outcomes of patients in colorectal cancer

Background: Colorectal cancer is the fourth most deadly cancer worldwide. Although current treatment regimens have prolonged the survival of patients, the prognosis is still unsatisfactory. Inflammation and lncRNAs are closely related to tumor occurrence and development in CRC. Therefore, it is necessary to establish a new prognostic signature based on inflammation-related lncRNAs to improve the prognosis of patients with CRC. Methods: LASSO-penalized Cox analysis was performed to construct a prognostic signature. Kaplan-Meier curves were used for survival analysis and ROC curves were used to measure the performance of the signature. Functional enrichment analysis was conducted to reveal the biological significance of the signature. The R package "maftool" and GISTIC2.0 algorithm were performed for analysis and visualization of genomic variations. The R package "pRRophetic", CMap analysis and submap analysis were performed to predict response to chemotherapy and immunotherapy. Results: An effective and independent prognostic signature, IRLncSig, was constructed based on sixteen inflammation-related lncRNAs. The IRLncSig was proved to be an independent prognostic indicator in CRC and was superior to clinical variables and the other four published signatures. The nomograms were constructed based on inflammation-related lncRNAs and detected by calibration curves. All samples were classified into two groups according to the median value, and we found frequent mutations of the TP53 gene in the high-risk group. We also found some significantly amplificated regions in the high-risk group, 8q24.3, 20q12, 8q22.3, and 20q13.2, which may regulate the inflammatory activity of cancer cells in CRC. Finally, we identified chemotherapeutic agents for high-risk patients and found that these patients were more likely to respond to immunotherapy, especially anti-CTLA4 therapy. Conclusion: In short, we constructed a new signature based on sixteen inflammation-related lncRNAs to improve the outcomes of patients in CRC. Our findings have proved that the IRLncSig can be used as an effective and independent marker for predicting the survival of patients with CRC.

A Gene Selection Method Based on Outliers for Breast Cancer Subtype Classification

Breast cancer is the second most common cancer type and is the leading cause of cancer-related deaths worldwide. Since it is a heterogeneous disease, subtyping breast cancer plays an important role in performing a specific treatment. Gene expression data is a viable alternative to be employed on cancer subtype classification, as they represent the state of a cell at the molecular level, but generally has a relatively small number of samples compared to a large number of genes. Gene selection is a promising approach that addresses this uneven high-dimensional matrix of genes versus samples and plays an important role in the development of efficient cancer subtype classification. In this work, an innovative outlier-based gene selection (OGS) method is proposed to select relevant genes for efficiently and effectively classify breast cancer subtypes. Experiments show that our strategy presents an F₁ score of 1.0 for basal and 0.86 for her 2, the two subtypes with the worst prognoses, respectively. Compared to other methods, our proposed method outperforms in the F₁ score using 80% less genes. In general, our method selects only a few highly relevant genes, speeding up the classification, and significantly improving the classifier's performance.

Identification of Prognostic Genes in Gliomas Based on Increased Microenvironment Stiffness

With a median survival time of 15 months, glioblastoma multiforme is one of the most aggressive primary brain cancers. The crucial roles played by the extracellular matrix (ECM) stiffness in glioma progression and treatment resistance have been reported in numerous studies. However, the association between ECM-stiffness-regulated genes and the prognosis of glioma patients remains to be explored. Thus, using bioinformatics analysis, we first identified 180 stiffness-dependent genes from an RNA-Seq dataset, and then evaluated their prognosis in The Cancer Genome Atlas (TCGA) glioma dataset. Our results showed that 11 stiffness-dependent genes common between low- and high-grade gliomas were prognostic. After validation using the Chinese Glioma Genome Atlas (CGGA) database, we further identified four stiffness-dependent prognostic genes: FN1, ITGA5, OSMR, and NGFR. In addition to high-grade glioma, overexpression of the four-gene signature also showed poor prognosis in low-grade glioma patients. Moreover, our analysis confirmed that the expression levels of stiffness-dependent prognostic genes in high-grade glioma were significantly higher than in low-grade glioma, suggesting that these genes were associated with glioma progression. Based on a pathophysiology-inspired approach, our findings illuminate the link between ECM stiffness and the prognosis of glioma patients and suggest a signature of four stiffness-dependent genes as potential therapeutic targets.

Identification of mRNA Signature for Predicting Prognosis Risk of Rectal Adenocarcinoma

Background: The immune system plays a crucial role in rectal adenocarcinoma (READ). Immune-related genes may help predict READ prognoses. Methods: The Cancer Genome Atlas dataset and GSE56699 were used as the training and validation datasets, respectively, and differentially expressed genes (DEGs) were identified. The optimal DEG combination was determined, and the prognostic risk model was constructed. The correlation between optimal DEGs and immune infiltrating cells was evaluated. Results: Nine DEGs were selected for analysis. Moreover, ADAMDEC1 showed a positive correlation with six immune infiltrates, most notably with B cells and dendritic cells. F13A1 was also positively correlated with six immune infiltrates, particularly macrophage and dendritic cells, whereas LGALS9C was negatively correlated with all immune infiltrates except B cells. Additionally, the prognostic risk model was strongly correlated with the actual situation. We retained only three prognosis risk factors: age, pathologic stage, and prognostic risk model. The stratified analysis revealed that lower ages and pathologic stages have a better prognosis with READ. Age and mRNA prognostic factors were the most important factors in determining the possibility of 3- and 5-year survival. Conclusion: In summary, we identified a nine-gene prognosis risk model that is applicable to the treatment of READ. Altogether, characteristics such as the gene signature and age have a strong predictive value for prognosis risk.

Identification of Hub Genes in Colorectal Adenocarcinoma by Integrated Bioinformatics

An improved understanding of the molecular mechanism of colorectal adenocarcinoma is necessary to predict the prognosis and develop new target gene therapy strategies. This study aims to identify hub genes associated with colorectal adenocarcinoma and further analyze their prognostic significance. In this study, The Cancer Genome Atlas (TCGA) COAD-READ database and the gene expression profiles of GSE25070 from the Gene Expression Omnibus were collected to explore the differentially expressed genes between colorectal adenocarcinoma and normal tissues. The weighted gene co-expression network analysis (WGCNA) and differential expression analysis identified 82 differentially co-expressed genes in the collected datasets. Enrichment analysis was applied to explore the regulated signaling pathway in colorectal adenocarcinoma. In addition, 10 hub genes were identified in the protein–protein interaction (PPI) network by using the cytoHubba plug-in of Cytoscape, where five genes were further proven to be significantly related to the survival rate. Compared with normal tissues, the expressions of the five genes were both downregulated in the GSE110224 dataset. Subsequently, the expression of the five hub genes was confirmed by the Human Protein Atlas database. Finally, we used Cox regression analysis to identify genes associated with prognosis, and a 3-gene signature (CLCA1–CLCA4–GUCA2A) was constructed to predict the prognosis of patients with colorectal cancer. In conclusion, our study revealed that the five hub genes and CLCA1–CLCA4–GUCA2A signature are highly correlated with the development of colorectal adenocarcinoma and can serve as promising prognosis factors to predict the overall survival rate of patients.

Predicting Colorectal Cancer Using Residual Deep Learning with Nursing Care

Presently, colorectal cancer is the second most dangerous cancer; around 13% of people have been affected; and it requires an effective image analysis and earlier cancer prediction (IAECP) system for reducing the mortality rate. Here, the IAECP system uses MRI radio imaging for predicting colorectal cancer. During this process, high- and low-level features are required to examine cancer in an earlier stage. Due to the limitation of the conventional feature extraction process, both features are difficult to extract from cancer suffered locations. Hence, a deep learning system (DLS) is used to examine the entire bowel MRI image to identify the cancer-affected location, feature extraction, and feature training process. Furthermore, the DLS-based IAECP system helps improve the overall colorectal cancer identification accuracy for further process. The derived bowel features are trained by applying the residual convolution network, which minimizes the error between predicted and actual values. Finally, the test query images are compared with the trained image by applying the sum, which is more absolute to the cross-correlation template feature matching (SACC) algorithm. The experimental process is performed using 100,000 histological data sets, which is considered a publicly available data set. Moreover, the introduced method does not use generic features, whereas the deep learning features help improve the overall IAECP prediction rate (99.8%) ratio as predicted at lab-scale analysis.

Apoptosis-Associated Gene Expression Profiling Is One New Prognosis Risk Predictor of Human Rectal Cancer

Background. Prior research has revealed the predictive significance of a series of genetic markers in the prognosis of rectal cancer (RC), but the roles of apoptosis-associated genes in RC are rarely studied. Methods. The RNA-seq data as well as clinical data about patients with rectum adenocarcinoma (READ) were downloaded from The Cancer Genome Atlas (TCGA) and the Genotype-Tissue Expression (GTEx) project. Additionally, 87 apoptosis-associated genes were downloaded and acquired from Kyoto Encyclopedia of Genes and Genomes (KEGG) database. Comprehensive bioinformatics analysis was carried out for deep exploration of the expression and prognostic significance of these genes. Least absolute shrinkage and selection operator (LASSO) and multivariate Cox regression analysis was performed for the establishment of a risk scoring equation for the prognosis model and construction of a survival prognosis model. ROC curves were drawn for evaluating the accuracy of the model. A real-time quantitative PCR assay was conducted for quantification of apoptosis-associated proteins related to prognosis. Results. Eight genes were identified as hub genes associated with the prognosis of PFS. A risk model of prognosis prediction based on four gene signatures (CYCS, IKBKB, NFKB1, and TRADD) was constructed. According to further analysis of this model, the high-risk group experienced worse overall survival than the other. The prognosis model demonstrated a favorable predictive ability, with areas under the receiver operating characteristic curves (AUC) of 0.720, 0.641, and 0.677 in forecasting the 1-, 2-, and 3-year prognosis, respectively. In addition, CYCS and NFKB1 presented low expression, while IKBKB and TRADD presented high expression in TCGA and clinical tumor samples. Conclusions. A four-gene signature risk model for prognosis forecasting of RC has been constructed, which possesses favorable predictive ability, which offers ideas and breakthrough points to the apoptosis-associated development of RC.

Identification of a 5-Nutrient Stress-Sensitive Gene Signature to Predict Survival for Colorectal Cancer

Background

The high heterogeneity and the complexity of the tumor microenvironment of colorectal cancer (CRC) have enhanced the difficulty of prognosis prediction based on conventional clinical indicators. Recent studies revealed that tumor cells could overcome various nutritional deficiencies by gene regulation and metabolic remodeling. However, whether differentially expressed genes (DEGs) in CRC cells under kinds of nutrient deficiency could be used to predict prognosis remained unveiled.

Methods

Three datasets (GSE70976, GSE13548, and GSE116087), in which colon cancer cells were, respectively, cultured in serum-free, glucose-free, or glutamine-free medium, were included to delineate the profiles of gene expression by nutrient stress. DEGs were figured out in three datasets, and gene functional analysis was performed. Survival analyses and Cox proportional hazards model were then used to identify nutrient stress sensitive genes in CRC datasets (GSE39582 and TCGA COAD). Then, a 5-gene signature was constructed and the risk scores were also calculated. Survival analyses, cox analyses, and nomogram were applied to predict the prognosis of patients with colorectal cancer. The effectiveness of the risk model was also tested.

Results

A total of 48 genes were found to be dysregulated in serum, glucose, or glutamine-deprived CRC cells, which were mainly enriched in cell cycle and endoplasmic reticulum stress pathways. After further analyses, 5 genes, MCM5, MCM6, CDCA2, GINS2, and SPC25, were identified to be differentially expressed in CRC and be related to prognosis of in CRC datasets. We used the above nutrient stress-sensitive genes to construct a risk scoring model. CRC samples in the datasets were divided into low-risk and high-risk groups. Data showed that higher risk scores were associated with better outcomes and risk scores decreased significantly with tumor procession. Moreover, the risk score could be used to predict the probability of survival based on nomogram.

Conclusions

The 5-nutrient stress-sensitive gene signature could act as an independent biomarker for survival prediction of CRC patients.

Bioinformatics analysis reveals immune prognostic markers for overall survival of colorectal cancer patients: a novel machine learning survival predictive system

Objectives

Immune microenvironment was closely related to the occurrence and progression of colorectal cancer (CRC). The objective of the current research was to develop and verify a Machine learning survival predictive system for CRC based on immune gene expression data and machine learning algorithms.

Methods

The current study performed differentially expressed analyses between normal tissues and tumor tissues. Univariate Cox regression was used to screen prognostic markers for CRC. Prognostic immune genes and transcription factors were used to construct an immune-related regulatory network. Three machine learning algorithms were used to create an Machine learning survival predictive system for CRC. Concordance indexes, calibration curves, and Brier scores were used to evaluate the performance of prognostic model.

Results

Twenty immune genes (BCL2L12, FKBP10, XKRX, WFS1, TESC, CCR7, SPACA3, LY6G6C, L1CAM, OSM, EXTL1, LY6D, FCRL5, MYEOV, FOXD1, REG3G, HAPLN1, MAOB, TNFSF11, and AMIGO3) were recognized as independent risk factors for CRC. A prognostic nomogram was developed based on the previous immune genes. Concordance indexes were 0.852, 0.778, and 0.818 for 1-, 3- and 5-year survival. This prognostic model could discriminate high risk patients with poor prognosis from low risk patients with favorable prognosis.

Conclusions

The current study identified twenty prognostic immune genes for CRC patients and constructed an immune-related regulatory network. Based on three machine learning algorithms, the current research provided three individual mortality predictive curves. The Machine learning survival predictive system was available at: https://zhangzhiqiao8.shinyapps.io/Artificial_Intelligence_Survival_Prediction_for_CRC_B1005_1/, which was valuable for individualized treatment decision before surgery.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12859-022-04657-3.

Identification of Ferroptosis-Related Genes Signature Predicting the Efficiency of Invasion and Metastasis Ability in Colon Adenocarcinoma

Background: Colon adenocarcinoma (COAD) is one of the most prevalent cancers worldwide and has become a leading cause of cancer death. Although many potential biomarkers of COAD have been screened with the bioinformatics method, it is necessary to explore novel markers for the diagnosis and appropriate individual treatments for COAD patients due to the high heterogeneity of this disease. Epithelial-to-mesenchymal transition (EMT)-mediated tumor metastasis suggests poor prognosis of cancers. Ferroptosis is involved in tumor development. EMT signaling can increase the cellular sensitivity to ferroptosis in tumors. The aim of our study is finding novel prognostic biomarkers to determine COAD patients for predicting efficiency of metastasis status and targeting precise ferroptosis-related therapy.

Methods: A novel gene signature related to metastasis and ferroptosis was identified combing with risk model and WGCNA analysis with R software. The biological functions and predictive ability of the signature in COAD were explored through bioinformatics analysis.

Results: We established a four-gene prognostic signature (MMP7, YAP1, PCOLCE, and HOXC11) based on EMT and ferroptosis related genes and validated the reliability and effectiveness of this model in COAD. This four-gene prognostic signature was closely connected with metastasis and ferroptosis sensitivity of COAD. Moreover, WGCNA analysis further confirmed the correlation between PCOLCE, HOXC11, and liver and lymphatic invasion of COAD.

Conclusion: The four genes may become potential prognostic biomarkers to identify COAD patients with metastasis. Moreover, this four-gene signature may be able to determine the COAD suitable with ferroptosis induction therapy. Finally, PCOLCE2 and HOXC11 were selected individually because of their novelties and precise prediction ability. Overall, this signature provided novel possibilities for better prognostic evaluation of COAD patients and may be of great guiding significance for individualized treatment and clinical decision.

A Novel DNA Damage Repair-Related Gene Signature for Predicting Glioma Prognosis

Background

Glioma is one of the most prevalent tumors in the central nervous system of adults and shows a poor prognosis. This study aimed to develop a DNA damage repair (DDR)-related gene signature to evaluate the prognosis of glioma patients.

Methods

Differentially expressed genes (DEGs) were extracted based on 276 DDR genes. Then, a gene signature was developed for the survival prediction in glioma patients by means of univariate, multivariate Cox, and least absolute shrinkage and selector operation (Lasso) analyses. After analyzing the clinical parameters, a nomogram was constructed and assessed. A total of 693 gliomas from the Chinese Glioma Genome Atlas (CGGA) were used for external validation. In addition, we used glioma tumor tissues for qPCR experiment to verify.

Results

A 12-DDR-related gene signature was identified from the 75 DEGs to stratify the survival risk of glioma patients. The overall survival of high-risk group was significantly shorter than that of low-risk group (P < 0.001). Besides, according to the risk score assessment, patients in high- or low-risk group also had significant correlations with clinicopathological parameters, including age (P < 0.01), grade (P < 0.001), IDH status (P < 0.001) and 1p19q codeletion status (P < 0.001). The nomogram provided favorable C-index and calibration plots. The C-index of training set and verification set was 0.761 and 0.746, respectively, and the calibration curve also showed that both training set and verification set were close to the standard curve. The qPCR results showed that there were significant differences in the expression of some typical DDR-related genes in tumor tissues and paracancer tissues (P_(WEE1)=0.0002, P_(RECQL)=0.0117, P_(RPA1)=0.021, P_(RRM1)=0.0035, P_(PARP4)=0.0006, P_(ELOA)=0.0023).

Conclusion

Our study developed a novel 12 DDR-related gene signature as a practical prognostic predictor for glioma patients. A nomogram combining the signature and clinical parameters was established as an individual clinical prediction tool.

A Network-Based Methodology to Identify Subnetwork Markers for Diagnosis and Prognosis of Colorectal Cancer

The development of reliable methods for identification of robust biomarkers for complex diseases is critical for disease diagnosis and prognosis efforts. Integrating multi-omics data with protein-protein interaction (PPI) networks to investigate diseases may help better understand disease characteristics at the molecular level. In this study, we developed and tested a novel network-based method to detect subnetwork markers for patients with colorectal cancer (CRC). We performed an integrated omics analysis using whole-genome gene expression profiling and copy number alterations (CNAs) datasets followed by building a gene interaction network for the significantly altered genes. We then clustered the constructed gene network into subnetworks and assigned a score for each significant subnetwork. We developed a support vector machine (SVM) classifier using these scores as feature values and tested the methodology in independent CRC transcriptomic datasets. The network analysis resulted in 15 subnetwork markers that revealed several hub genes that may play a significant role in colorectal cancer, including PTP4A3, FGFR2, PTX3, AURKA, FEN1, INHBA, and YES1. The 15-subnetwork classifier displayed over 98 percent accuracy in detecting patients with CRC. In comparison to individual gene biomarkers, subnetwork markers based on integrated multi-omics and network analyses may lead to better disease classification, diagnosis, and prognosis.

Comprehensive analysis of immune related lncRNAs in the tumor microenvironment of stage II-III colorectal cancer

Background

Long non-coding RNAs (lncRNAs) associated with immunological function have increasingly been found to act as effective prognostic biomarkers of the overall survival (OS) of colorectal cancer (CRC) patients. We sought to identify a signature of immune-related lncRNAs that offered value as a tool for the prospective prognostic evaluation of patients with stage II-III CRC.

Methods

The clinical and gene expression data of CRC patients in The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) databases was obtained and separated into a training cohort composed of 202 samples, a test cohort of 124 samples from the GSE72970 dataset, and a validation cohort of 91 samples from the GSE143985 dataset.

Results

We firstly evaluated intratumoral immune cell infiltration by conducting a Single-sample gene set enrichment analyses (ssGSEA) analysis to separate patient tumors into those with low immune cell infiltration and those with high immune cell infiltration. We then compared lncRNA and mRNA expression profiles between these two tumor types, leading us to focus on eight lncRNAs identified within the resultant mRNA-lncRNA co-expression network. Multivariate Cox regression models were then utilized to detect an immune-associated lncRNA signature that offered value for prognostic model construction. Functional analyses revealed this lncRNA signature to be associated with key immunological pathways including the JAK-STAT signaling, T cell receptor signaling, and Rap1 signaling pathways.

Conclusions

Together, our results suggest that our immune-related 4 lncRNA signature can reliably predict stage II-III CRC patient prognosis, thereby guiding efforts to better understand this disease and to effectively treat it.

Integrative Transcriptome Profiling Reveals SKA3 as a Novel Prognostic Marker in Non-Muscle Invasive Bladder Cancer

Approximately 80% of all new bladder cancer patients are diagnosed with non-muscle invasive bladder cancer (NMIBC). However, approximately 15% of them progress to muscle-invasive bladder cancer (MIBC), for which prognosis is poor. The current study aimed to improve diagnostic accuracy associated with clinical outcomes in NMIBC patients. Nevertheless, it has been challenging to identify molecular biomarkers that accurately predict MIBC progression because this disease is complex and heterogeneous. Through integrative transcriptome profiling, we showed that high SKA3 expression is associated with poor clinical outcomes and MIBC progression. We performed RNA sequencing on human tumor tissues to identify candidate biomarkers in NMIBC. We then selected genes with prognostic significance by analyzing public datasets from multiple cohorts of bladder cancer patients. We found that SKA3 was associated with NMIBC pathophysiology and poor survival. We analyzed public single-cell RNA-sequencing (scRNA-seq) data for bladder cancer to dissect transcriptional tumor heterogeneity. SKA3 was expressed in an epithelial cell subpopulation expressing genes regulating the cell cycle. Knockdown experiments confirmed that SKA3 promotes bladder cancer cell proliferation by accelerating G2/M transition. Hence, SKA3 is a new prognostic marker for predicting NMIBC progression. Its inhibition could form part of a novel treatment lowering the probability of bladder cancer progression.

The role of multiple metabolic genes in predicting the overall survival of colorectal cancer: A study based on TCGA and GEO databases

The recent advances in gene chip technology have led to the identification of multiple metabolism-related genes that are closely associated with colorectal cancer (CRC). Nevertheless, none of these genes could accurately diagnose or predict CRC. The prognosis of CRC has been made by previous prognostic models constructed by using multiple genes, however, the predictive function of multi-gene prognostic models using metabolic genes for the CRC prognosis remains unexplored. In this study, we used the TCGA-CRC cohort as the test dataset and the GSE39582 cohort as the experimental dataset. Firstly, we constructed a prognostic model using metabolic genes from the TCGA-CRC cohort, which were also associated with CRC prognosis. We analyzed the advantages of the prognostic model in the prognosis of CRC and its regulatory mechanism of the genes associated with the model. Secondly, the outcome of the TCGA-CRC cohort analysis was validated using the GSE39582 cohort. We found that the prognostic model can be employed as an independent prognostic risk factor for estimating the CRC survival rate. Besides, compared with traditional clinical pathology, it can precisely predict CRC prognosis as well. The high-risk group of the prognostic model showed a substantially lower survival rate as compared to the low-risk group. In addition, gene enrichment analysis of metabolic genes showed that genes in the prognostic model are enriched in metabolism and cancer-related pathways, which may explain its underlying mechanism. Our study identified a novel metabolic profile containing 11 genes for prognostic prediction of CRC. The prognostic model may unravel the imbalanced metabolic microenvironment, and it might promote the development of biomarkers for predicting treatment response and streamlining metabolic therapy in CRC.

Weighted Gene Co-expression Network Analysis Identified a Novel Thirteen-Gene Signature Associated With Progression, Prognosis, and Immune Microenvironment of Colon Adenocarcinoma Patients

Colon adenocarcinoma (COAD) is one of the most common malignant tumors and has high migration and invasion capacity. In this study, we attempted to establish a multigene signature for predicting the prognosis of COAD patients. Weighted gene co-expression network analysis and differential gene expression analysis methods were first applied to identify differentially co-expressed genes between COAD tissues and normal tissues from the Cancer Genome Atlas (TCGA)-COAD dataset and GSE39582 dataset, and a total of 309 overlapping genes were screened out. Then, our study employed TCGA-COAD cohort as the training dataset and an independent cohort by merging the GES39582 and GSE17536 datasets as the testing dataset. After univariate and multivariate Cox regression analyses were performed for these overlapping genes and overall survival (OS) of COAD patients in the training dataset, a 13-gene signature was constructed to divide COAD patients into high- and low-risk subgroups with significantly different OS. The testing dataset exhibited the same results utilizing the same predictive signature. The area under the curve of receiver operating characteristic analysis for predicting OS in the training and testing datasets were 0.789 and 0.868, respectively, which revealed the enhanced predictive power of the signature. Multivariate Cox regression analysis further suggested that the 13-gene signature could independently predict OS. Among the 13 prognostic genes, NAT1 and NAT2 were downregulated with deep deletions in tumor tissues in multiple COAD cohorts and exhibited significant correlations with poorer OS based on the GEPIA database. Notably, NAT1 and NAT2 expression levels were positively correlated with infiltrating levels of CD8+ T cells and dendritic cells, exhibiting a foundation for further research investigating the antitumor immune roles played by NAT1 and NAT2 in COAD. Taken together, the results of our study showed that the 13-gene signature could efficiently predict OS and that NAT1 and NAT2 could function as biomarkers for prognosis and the immune response in COAD.

Identification and validation of a novel glycolysis-related gene signature for predicting the prognosis in ovarian cancer

Background

Ovarian cancer (OC) is the most lethal gynaecological tumor. Changes in glycolysis have been proven to play an important role in OC progression. We aimed to identify a novel glycolysis-related gene signature to better predict the prognosis of patients with OC.

Methods

mRNA and clinical data were obtained from The Cancer Genome Atlas (TCGA), International Cancer Genome Consortium (ICGC) and Genotype Tissue Expression (GTEx) database. The “limma” R package was used to identify glycolysis-related differentially expressed genes (DEGs). Then, a multivariate Cox proportional regression model and survival analysis were used to develop a glycolysis-related gene signature. Furthermore, the TCGA training set was divided into two internal test sets for validation, while the ICGC dataset was used as an external test set. A nomogram was constructed in the training set, and the relative proportions of 22 types of tumor-infiltrating immune cells were evaluated using the “CIBERSORT” R package. The enriched Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways were determined by single-sample gene set enrichment analysis (ssGSEA) with the “GSVA” R package. Finally, the expression and function of the unreported signature genes ISG20 and SEH1L were explored using immunohistochemistry, western blotting, qRT-PCR, proliferation, migration, invasion and xenograft tumor assays.

Results

A five-gene signature comprising ANGPTL4, PYGB, ISG20, SEH1L and IRS2 was constructed. This signature could predict prognosis independent of clinical factors. A nomogram incorporating the signature and three clinical features was constructed, and the calibration plot suggested that the nomogram could accurately predict the survival rate. According to ssGSEA, the signature was associated with KEGG pathways related to axon guidance, mTOR signalling, tight junctions, etc. The proportions of tumor-infiltrating immune cells differed significantly between the high-risk group and the low-risk group. The expression levels of ISG20 and SEH1L were lower in tumor tissues than in normal tissues. Overexpression of ISG20 or SEH1L suppressed the proliferation, migration and invasion of Caov3 cells in vitro and the growth of xenograft tumors in vivo.

Conclusion

Five glycolysis-related genes were identified and incorporated into a novel risk signature that can effectively assess the prognosis and guide the treatment of OC patients.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12935-021-02045-0.

Construction and validation of a metabolic risk model predicting prognosis of colon cancer

Metabolic genes have played a significant role in tumor development and prognosis. In this study, we constructed a metabolic risk model to predict the prognosis of colon cancer based on The Cancer Genome Atlas (TCGA) and validated the model by Gene Expression Omnibus (GEO). We extracted 753 metabolic genes and identified 139 differentially expressed genes (DEGs) from TCGA database. Then we conducted univariate cox regression analysis and Least Absolute Shrinkage and Selection Operator Cox regression analysis to identify prognosis-related genes and construct the metabolic risk model. An eleven-gene prognostic model was constructed after 1000 resamples. The gene signature has been proved to have an excellent ability to predict prognosis by Kaplan-Meier analysis, time-dependent receiver operating characteristic, risk score, univariate and multivariate cox regression analysis based on TCGA. Then we validated the model by Kaplan-Meier analysis and risk score based on GEO database. Finally, we performed a weighted gene co-expression network analysis and protein-protein interaction network on DEGs, and Kyoto Encyclopedia of Genes and Genomes pathways and Gene Ontology enrichment analyses were conducted. The results of functional analyses showed that most significantly enriched pathways focused on metabolism, especially glucose and lipid metabolism pathways.

Identification of a 5-Gene-Based Scoring System by WGCNA and LASSO to Predict Prognosis for Rectal Cancer Patients

Background

Although accumulating evidence suggested that a molecular signature panel may be more effective for the prognosis prediction than routine clinical characteristics, current studies mainly focused on colorectal or colon cancers. No reports specifically focused on the signature panel for rectal cancers (RC). Our present study was aimed at developing a novel prognostic signature panel for RC.

Methods

Sequencing (or microarray) data and clinicopathological details of patients with RC were retrieved from The Cancer Genome Atlas (TCGA-READ) or the Gene Expression Omnibus (GSE123390, GSE56699) database. A weighted gene coexpression network was used to identify RC-related modules. The least absolute shrinkage and selection operator analysis was performed to screen the prognostic signature panel. The prognostic performance of the risk score was evaluated by survival curve analyses. Functions of prognostic genes were predicted based on the interaction proteins and the correlation with tumor-infiltrating immune cells. The Human Protein Atlas (HPA) tool was utilized to validate the protein expression levels.

Results

A total of 247 differentially expressed genes (DEGs) were commonly identified using TCGA and GSE123390 datasets. Brown and yellow modules (including 77 DEGs) were identified to be preserved for RC. Five DEGs (ASB2, GPR15, PRPH, RNASE7, and TCL1A) in these two modules constituted the optimal prognosis signature panel. Kaplan-Meier curve analysis showed that patients in the high-risk group had a poorer prognosis than those in the low-risk group. Receiver operating characteristic (ROC) curve analysis demonstrated that this risk score had high predictive accuracy for unfavorable prognosis, with the area under the ROC curve of 0.915 and 0.827 for TCGA and GSE56699 datasets, respectively. This five-mRNA classifier was an independent prognostic factor. Its predictive accuracy was also higher than all clinical factor models. A prognostic nomogram was developed by integrating the risk score and clinical factors, which showed the highest prognostic power. ASB2, PRPH, and GPR15/TCL1A were predicted to function by interacting with CASQ2/PDK4/EPHA67, PTN, and CXCL12, respectively. TCL1A and GPR15 influenced the infiltration levels of B cells and dendritic cells, while the expression of PRPH was positively associated with the abundance of macrophages. HPA analysis supported the downregulation of PRPH, RNASE7, CASQ2, EPHA6, and PDK4 in RC compared with normal controls.

Conclusion

Our immune-related signature panel may be a promising prognostic indicator for RC.

Development of a Gene-Based Prediction Model for Recurrence of Colorectal Cancer Using an Ensemble Learning Algorithm

It is difficult to determine which patients with stage I and II colorectal cancer are at high risk of recurrence, qualifying them to undergo adjuvant chemotherapy. In this study, we aimed to determine a gene signature using gene expression data that could successfully identify high risk of recurrence among stage I and II colorectal cancer patients. First, a synthetic minority oversampling technique was used to address the problem of imbalanced data due to rare recurrence events. We then applied a sequential workflow of three methods (significance analysis of microarrays, logistic regression, and recursive feature elimination) to identify genes differentially expressed between patients with and without recurrence. To stabilize the prediction algorithm, we repeated the above processes on 10 subsets by bagging the training data set and then used support vector machine methods to construct the prediction models. The final predictions were determined by majority voting. The 10 models, using 51 differentially expressed genes, successfully predicted a high risk of recurrence within 3 years in the training data set, with a sensitivity of 91.18%. For the validation data sets, the sensitivity of the prediction with samples from two other countries was 80.00% and 91.67%. These prediction models can potentially function as a tool to decide if adjuvant chemotherapy should be administered after surgery for patients with stage I and II colorectal cancer.

Identification and Validation of a Novel Six-Gene Prognostic Signature of Stem Cell Characteristic in Colon Cancer

Cancer stem cells play crucial roles in the development of colon cancer (COAD). This study tried to explore new markers for predicting the prognosis of colon cancer based on stem cell-related genes. In our study, 424 COAD samples from TCGA were divided into three subtypes based on 412 stem cell-related genes; there were significant differences in prognosis, clinical characteristics, and immune scores between these subtypes. 694 genes were screened between subgroups. Subsequently a six-gene signature (DYDC2, MS4A15, MAGEA1, WNT7A, APOD, and SERPINE1) was established. This model had strong robustness and stable predictive performance in cohorts of different platforms. Taken together, the six-gene signature constructed in this study could be used as a novel prognostic marker for COAD patients.

Identification of DNA repair-related genes predicting pathogenesis and prognosis for liver cancer

Background

Liver cancer (LC) is one of the most fatal cancers throughout the world. More efficient and sensitive gene signatures that could accurately predict survival in LC patients are vitally needed to promote a better individualized and effective treatment.

Material/methods

422 LC and adjacent normal tissues with both RNA-Seq and clinical data in TCGA were embedded in our study. Gene set enrichment analysis (GSEA) was applied to identify genes and hallmark gene sets that are more valuable for liver cancer therapy. Cox regression analysis was used to identify genes related to overall survival (OS) and build the prediction model. cBioPortal database was used to examine the alterations of the panel mRNA signature. ROC curves and Kaplan–Meier curves were used to validate the prediction model. Besides, the expression of the genes in the model were validated using quantitative real-time PCR in clinical tissue specimens.

Results

The panel of DNA repair-related mRNA signature consisted of seven mRNAs: RFC4 (replication factor C subunit 4), ZWINT (ZW10 interacting kinetochore protein), UPF3B (UPF3B regulator of nonsense mediated mRNA decay), NCBP2 (nuclear cap binding protein subunit 2), ADA (adenosine deaminase), SF3A3 (splicing factor 3a subunit 3) and GTF2H1 (general transcription factor IIH subunit 1). On-line analysis of cBioPortal database found that the expression of the panel mRNA has a wide variation ranging from 7 to 10%. All the mRNAs were significantly upregulated in LC tissues compared to normal tissues (P < 0.05). The risk model is closely related to the OS of LC patients. The hazard ratio (HR) is 2.184 [95% CI (confidence interval) 1.523–3.132] and log-rank P-value < 0.0001. For clinical specimen validation, we found that all of the genes in the model upregulated in liver cancer tissues versus normal liver tissues, which was consistent with the results predicted.

Conclusions

Our study demonstrated a mRNA signature including seven mRNA for prognosis prediction of LC. This panel gene signature provides a new criterion for accurate diagnosis and therapeutic target of LC.

Construction of a Prognostic Risk Prediction Model for Obesity Combined With Breast Cancer

The improvement in the quality of life is accompanied by an accelerated pace of living and increased work-related pressures. Recent decades has seen an increase in the proportion of obese patients, as well as an increase in the prevalence of breast cancer. More and more evidences prove that obesity may be one of a prognostic impact factor in patients with breast cancer. Obesity presents unique diagnostic and therapeutic challenges in the population of breast cancer patients. Therefore, it is essential to have a better understanding of the relationship between obesity and breast cancer. This study aims to construct a prognostic risk prediction model combining obesity and breast cancer. In this study, we obtained a breast cancer sample dataset from the GEO database containing obesity data [determined by the body mass index (BMI)]. A total of 1174 genes that were differentially expressed between breast cancer samples of patients with and without obesity were screened by the rank-sum test. After weighted gene co-expression network analysis (WGCNA), 791 related genes were further screened. Relying on single-factor COX regression analysis to screen the candidate genes to 30, these 30 genes and another set of TCGA data were intersected to obtain 24 common genes. Finally, lasso regression analysis was performed on 24 genes, and a breast cancer prognostic risk prediction model containing 6 related genes was obtained. The model was also found to be related to the infiltration of immune cells. This study provides a new and accurate prognostic model for predicting the survival of breast cancer patients with obesity.

Identification of an Immune-Related Gene Signature to Improve Prognosis Prediction in Colorectal Cancer Patients

BACKGROUND

Despite recent advance in immune therapy, great heterogeneity exists in the outcomes of colorectal cancer (CRC) patients. In this study, we aimed to analyze the immune-related gene (IRG) expression profiles from three independent public databases and develop an effective signature to forecast patient's prognosis.

METHODS

IRGs were collected from the ImmPort database. The CRC dataset from The Cancer Genome Atlas (TCGA) database was used to identify a prognostic gene signature, which was verified in another two CRC datasets from the Gene Expression Omnibus (GEO). Gene function enrichment analysis was conducted. A prognostic nomogram was built incorporating the IRG signature with clinical risk factors.

RESULTS

The three datasets had 487, 579, and 224 patients, respectively. A prognostic six-gene-signature (CCL22, LIMK1, MAPKAPK3, FLOT1, GPRC5B, and IL20RB) was developed through feature selection that showed good differentiation between the low- and high-risk groups in the training set (p < 0.001), which was later confirmed in the two validation groups (log-rank p < 0.05). The signature outperformed tumor TNM staging for survival prediction. GO and KEGG functional annotation analysis suggested that the signature was significantly enriched in metabolic processes and regulation of immunity (p < 0.05). When combined with clinical risk factors, the model showed robust prediction capability.

CONCLUSION

The immune-related six-gene signature is a reliable prognostic indicator for CRC patients and could provide insight for personalized cancer management.

Prognostic Model of Colorectal Cancer Constructed by Eight Immune-Related Genes

Background

Colorectal cancer (CRC) is a common malignant tumor of the digestive tract with a high mortality rate. Growing evidence demonstrates that immune-related genes play a prominent role in the occurrence and development of CRC. The aim of this study was to investigate the prognostic value of immune-related genes in CRC.

Methods

Gene expression profiles and clinical data of 568 CRC and 44 non-tumorous tissues were obtained from The Cancer Genome Atlas (TCGA) database. First, we performed a differentially expressed gene (DEG) analysis and univariate Cox regression analysis to determine the DEGs associated with overall survival. Gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses were subsequently performed for prognostic immune-related genes. Then, a multivariate Cox regression analysis was performed to establish the immune prognostic model and identify the independent prognostic factors of CRC. Next, in vitro experiments were done to further validate the model. Finally, we analyzed the correlation among immune-related genes, clinical traits, and immune cell infiltration.

Results

In total, 3,702 DEGs were obtained, and 338 prognostic immune-related genes were identified. Among them, 45 genes were significantly correlated with the prognosis of CRC patients. A TF-mediated network was set up to explore its internal mechanism. GO and KEGG analyses further illustrated that these genes were enriched in immune-and inflammatory-related pathways. Then, a prognostic prediction model composed of eight immune-related genes (SLC10A2, UTS2, FGF2, UCN, IL1RL2, ESM1, ADIPOQ, and VIP) was constructed. The AUC of the ROC curve for 1, 3, 5, and 10 years overall survival (OS) was 0.751, 0.707, 0.680, and 0.729, respectively. The survival analysis suggested that the OS of the high-risk group was significantly poorer than that of the low-risk group. Meanwhile, in vitro assays revealed that ESM1 and SLC10A2 exert opposing roles in colon cancer cell proliferation, validating the accuracy of the model. The correlation analysis indicated that immune cell infiltration was positively related to the model.

Conclusion

This study screened prognosis-related immune genes and developed a prognostic prediction model of CRC. These findings may help provide potential novel prognostic biomarkers and therapeutic targets for CRC. At the same time, the understanding of the CRC immune microenvironment status was deepened.

Gene Expression Along with Genomic Copy Number Variation and Mutational Analysis Were Used to Develop a 9-Gene Signature for Estimating Prognosis of COAD

PURPOSE

This study aims to systematically analyze multi-omics data to explore new prognosis biomarkers in colon adenocarcinoma (COAD).

MATERIALS AND METHODS

Multi-omics data of COAD and clinical information were obtained from The Cancer Genome Atlas (TCGA). Univariate Cox analysis was used to select genes which significantly related to the overall survival. GISTIC 2.0 software was used to identify significant amplification or deletion. Mutsig 2.0 software was used to identify significant mutation genes. The 9-gene signature was screened by random forest algorithm and Cox regression analysis. GSE17538 dataset was used as an external dataset to verify the predictive ability of 9-gene signature. qPCR was used to detect the expression of 9 genes in clinical specimens.

RESULTS

A total of 71 candidate genes are obtained by integrating genomic variation, mutation and prognostic data. Then, 9-gene signature was established, which includes HOXD12, RNF25, CBLN3, DOCK3, DNAJB13, PYGO2, CTNNA1, PTPRK, and NAT1. The 9-gene signature is an independent prognostic risk factor for COAD patients. In addition, the signature shows good predicting performance and clinical practicality in training set, testing set and external verification set. The results of qPCR based on clinical samples showed that the expression of HOXD12, RNF25, CBLN3, DOCK3, DNAJB13, and PYGO2 was increased in colon cancer tissues and the expression of CTNNA1, PTPRK, NAT1 was decreased in colon cancer tissues.

CONCLUSION

In this study, 9-gene signature is constructed as a new prognostic marker to predict the survival of COAD patients.

Identification and verification of three key genes associated with survival and prognosis of COAD patients via integrated bioinformatics analysis

BACKGROUND

Colorectal cancer (CRC) is the third most lethal malignancy in the world, wherein colon adenocarcinoma (COAD) is the most prevalent type of CRC. Exploring biomarkers is important for the diagnosis, treatment, and prevention of COAD.

METHODS

We used GEO2R and Venn online software for differential gene screening analysis. Hub genes were screened via Search Tool for the Retrieval of Interacting Genes/Proteins (STRING) and Cytoscape, following Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis. Finally, survival analysis and RNA expression validation were performed via UALCAN online software and real-time PCR. Immunohistochemistry (IHC) was performed to verify the protein expression level of hub genes from tissues of COAD patients.

RESULTS

In the present study, we screened 323 common differentially expressed genes (DEGs) from four GSE datasets. Furthermore, four hub genes were selected for survival correlation analysis and expression level verification, three of which were shown to be statistically significant.

CONCLUSION

Our study suggests that Serpin Family E Member 1 (SERPINE1), secreted phosphoprotein 1 (SPP1) and tissue inhibitor of metalloproteinase 1 (TIMP1) may be biomarkers closely related to the prognosis of CRC patients.

Prognostic Value of SLC4A4 and its Correlation with Immune Infiltration in Colon Adenocarcinoma

BACKGROUND SLC4A4 is differentially expressed in a variety of tumors, but its significance in colon adenocarcinoma has not been determined. MATERIAL AND METHODS Transcriptomes of two cohorts, GSE41258 and GSE32323, contained in The Cancer Genome Atlas (TCGA) were analysed to determine differences in SLC4A4 expression between tumor and normal tissue and their correlations with overall survival. The relationships between SLC4A4 expression and clinical characteristics were determined by COX regression analysis and logistic regression analysis, and correlations of SLC4A4 levels with tumor infiltrating immune cells (TIICs) and genes with high mutation frequency were evaluated by Pearson correlation analysis. Molecular functions and signaling pathways that might be affected by changes in SLC4A4 expression were determined by gene set enrichment analysis (GSEA). The overall distribution of TIICs was determined by two web servers: tumor immune estimation resource (TIMER) and CIBERSORT. RESULTS SLC4A4 expression was lower in colon adenocarcinoma than in normal colon tissue, suggesting that SLC4A4 was associated with poor prognosis. Reduced SLC4A4 expression was also associated with lymph node invasion and distant metastasis and was moderately correlated with increased expression of MUC4 and SMAD4, two genes with high mutation frequency in colon adenocarcinoma. GSEA indicated that changes in SLC4A4 expression affects several biological processes, including mismatch repair, base excision repair, and DNA replication. Eight TIICs in the tumor microenvironment differed significantly in groups with low and high expression of SLC4A4. CONCLUSIONS SLC4A4 may be a novel biomarker predicting prognosis in patients with colon adenocarcinoma. TIICs differed significantly in samples with higher and lower expression of SLC4A4.

Identification of a nine-gene prognostic signature for gastric carcinoma using integrated bioinformatics analyses

BACKGROUND

Gastric carcinoma (GC) is one of the most aggressive primary digestive cancers. It has unsatisfactory therapeutic outcomes and is difficult to diagnose early.

AIM

To identify prognostic biomarkers for GC patients using comprehensive bioinformatics analyses.

METHODS

Differentially expressed genes (DEGs) were screened using gene expression data from The Cancer Genome Atlas and Gene Expression Omnibus databases for GC. Overlapping DEGs were analyzed using univariate and multivariate Cox regression analyses. A risk score model was then constructed and its prognostic value was validated utilizing an independent Gene Expression Omnibus dataset (GSE15459). Multiple databases were used to analyze each gene in the risk score model. High-risk score-associated pathways and therapeutic small molecule drugs were analyzed and predicted, respectively.

RESULTS

A total of 95 overlapping DEGs were found and a nine-gene signature (COL8A1, CTHRC1, COL5A2, AADAC, MAMDC2, SERPINE1, MAOA, COL1A2, and FNDC1) was constructed for the GC prognosis prediction. Receiver operating characteristic curve performance in the training dataset (The Cancer Genome Atlas-stomach adenocarcinoma) and validation dataset (GSE15459) demonstrated a robust prognostic value of the risk score model. Multiple database analyses for each gene provided evidence to further understand the nine-gene signature. Gene set enrichment analysis showed that the high-risk group was enriched in multiple cancer-related pathways. Moreover, several new small molecule drugs for potential treatment of GC were identified.

CONCLUSION

The nine-gene signature-derived risk score allows to predict GC prognosis and might prove useful for guiding therapeutic strategies for GC patients.

Gene signature and prognostic merit of M6a regulators in colorectal cancer

IMPACT STATEMENT

Although new diagnostic techniques and treatments are increasingly updated for CRC, the clinical outcomes of CRC patients are still not encouraging with a low survival rate. N6-methyladenosine (m6A) as a popular modification on mRNA is associated with multiple types of cancers. Our purpose is to identify gene signature and prognostic ability of m6A modulators in CRC. For the first time, we identified genetic changes of m6A modulators and built prognostic gene signature in CRC, which may provide effective targets for the diagnosis and management of CRC.

Identification of Hub Genes Related to Carcinogenesis and Prognosis in Colorectal Cancer Based on Integrated Bioinformatics

The high mortality of colorectal cancer (CRC) patients and the limitations of conventional tumor-node-metastasis (TNM) stage emphasized the necessity of exploring hub genes closely related to carcinogenesis and prognosis in CRC. The study is aimed at identifying hub genes associated with carcinogenesis and prognosis for CRC. We identified and validated 212 differentially expressed genes (DEGs) from six Gene Expression Omnibus (GEO) datasets and the Cancer Genome Atlas (TCGA) database. We investigated functional enrichment analysis for DEGs. The protein-protein interaction (PPI) network was constructed, and hub modules and genes in CRC carcinogenesis were extracted. A prognostic signature was developed and validated based on Cox proportional hazards regression analysis. The DEGs mainly regulated biological processes covering response to stimulus, metabolic process, and affected molecular functions containing protein binding and catalytic activity. The DEGs played important roles in CRC-related pathways involving in preneoplastic lesions, carcinogenesis, metastasis, and poor prognosis. Hub genes closely related to CRC carcinogenesis were extracted including six genes in model 1 (CXCL1, CXCL3, CXCL8, CXCL11, NMU, and PPBP) and two genes and Metallothioneins (MTs) in model 2 (SLC26A3 and SLC30A10). Among them, CXCL8 was also related to prognosis. An eight-gene signature was proposed comprising AMH, WBSCR28, SFTA2, MYH2, POU4F1, SIX4, PGPEP1L, and PAX5. The study identified hub genes in CRC carcinogenesis and proposed an eight-gene signature with good reproducibility and robustness at the molecular level for CRC, which might provide directive significance for treatment selection and survival prediction.

A Nine-Gene Signature for Predicting the Response to Preoperative Chemoradiotherapy in Patients with Locally Advanced Rectal Cancer

Preoperative chemoradiotherapy (PCRT) and subsequent surgery is the standard multimodal treatment for locally advanced rectal cancer (LARC), albeit PCRT response varies among the individuals. This creates a dire necessity to identify a predictive model to forecast treatment response outcomes and identify patients who would benefit from PCRT. In this study, we performed a gene expression study using formalin-fixed paraffin-embedded (FFPE) tumor biopsy samples from 156 LARC patients (training cohort n = 60; validation cohort n = 96); we identified the nine-gene signature (FGFR3, GNA11, H3F3A, IL12A, IL1R1, IL2RB, NKD1, SGK2, and SPRY2) that distinctively differentiated responders from non-responders in the training cohort (accuracy = 86.9%, specificity = 84.8%, sensitivity = 81.5%) as well as in an independent validation cohort (accuracy = 81.0%, specificity = 79.4%, sensitivity = 82.3%). The signature was independent of all pathological and clinical features and was robust in predicting PCRT response. It is readily applicable to the clinical setting using FFPE samples and Food and Drug Administration (FDA) approved hardware and reagents. Predicting the response to PCRT may aid in tailored therapies for respective responders to PCRT and improve the oncologic outcomes for LARC patients.

Network Properties of Cancer Prognostic Gene Signatures in the Human Protein Interactome

Prognostic gene signatures are critical in cancer prognosis assessments and their pinpoint treatments. However, their network properties remain unclear. Here, we obtained nine prognostic gene sets including 1439 prognostic genes of different cancers from related publications. Four network centralities were used to examine the network properties of prognostic genes (PG) compared with other gene sets based on the Human Protein Reference Database (HPRD) and String networks. We also proposed three novel network measures for further investigating the network properties of prognostic gene sets (PGS) besides clustering coefficient. The results showed that PG did not occupy key positions in the human protein interaction network and were more similar to essential genes rather than cancer genes. However, PGS had significantly smaller intra-set distance (IAD) and inter-set distance (IED) in comparison with random sets (p-value < 0.001). Moreover, we also found that PGS tended to be distributed within network modules rather than between modules (p-value < 0.01), and the functional intersection of the modules enriched with PGS was closely related to cancer development and progression. Our research reveals the common network properties of cancer prognostic gene signatures in the human protein interactome. We argue that these are biologically meaningful and useful for understanding their molecular mechanism.

A Combined four-mRNA Signature Associated with Lymphatic Metastasis for Prognosis of Colorectal Cancer

Background: Colorectal cancer (CRC) is one of the most common malignant tumors in the world. Lymph node metastasis (LNM) is a common mode of metastasis of CRC. However, the combined mRNA biomarkers associated with LNM of CRC that can effectively predict CRC prognosis have not been reported yet.

Methods: To identify biomarkers that are associated with LNM, we collected data from the The Cancer Genome Atlas (TCGA) database. The edgeR package was searched to seek LNM-related genes by comparisons between cancer samples and normal colorectal tissues and between LNM and non-LNM (NLNM) of CRC. Univariate and multivariate regression analysis of genes in the intersection to build gene signature associated with independent prognosis of CRC, and then verified by Kaplan-Meier curve and log-rank test, receiver operating characteristic (ROC) curve was used to determine the efficiency of survival prediction of our four-mRNA signature. Finally, the potential molecular mechanisms and properties of these gene signature were also explored with functional and pathway enrichment analysis.

Results: 329 mRNAs were up-regulated in CRC tissues with LNM, and 8461 mRNAs were up-regulated in CRC tissues, the intersection is 100 mRNAs. After univariate and multivariate Cox regression analysis of 100 mRNAs, a novel four LNM related mRNAs (EPHA8, KRT85, GABRA3, and CLPSL1) were screened as independent prognostic indicators of CRC. Surprisingly, the four-mRNA signature can predict the prognosis of CRC patients independently of clinical factors andthe area under the curve (AUC) of the ROC is 0.730. The novel four-mRNA signature was used to identify high and low-risk groups. Stratified analysis indicated the risk score based on four-mRNA signature was an independent prognostic indicator for female, T3+T4, N1+N2 ,stage III+IV and patients with no new tumor event. Functional annotation of this risk model in high-risk patients revealed that pathways associated with neuroactive ligand-receptor interaction, estrogen signaling pathway, and steroid hormone biosynthesis.

Conclusions: By conducting TCGA data mining, our study demonstrated that a four-mRNA signature associated with LNM can be used as a combined biomarker for independent prognosis of CRC.

Network-based integration method for potential breast cancer gene identification

Breast cancer is the most common female death-causing cancer worldwide. A network-based integration method was proposed to identify potential breast cancer genes. First, genes were prioritized using a gene prioritization algorithm by the strategy of disease risks transferred between genes in a network with weighted vertexes and edges. Our prioritization algorithm was effectives and robust for top-ranked seed gene number and higher area under the curve values compared to ToppGene and ToppNet. Then, 20 potential breast cancer genes were identified as common genes of the top 50 candidate genes for their robustness in multiple prioritizations. These genes could accurately classify tumor and normal samples of all and paired sample sets and three independent datasets. Of potential breast cancer genes, 18 were verified by literature and 2 were novel genes that need further study. This study would contribute to the understanding of the genetic architecture for the diagnosis and treatment of breast cancer.

Identification of a Prognostic 3-Gene Risk Prediction Model for Thyroid Cancer

Objective: We aimed to screen the genes associated with thyroid cancer (THCA) prognosis, and construct a poly-gene risk prediction model for prognosis prediction and improvement. Methods: The HTSeq-Counts data of THCA were accessed from TCGA database, including 505 cancer samples and 57 normal tissue samples. "edgeR" package was utilized to perform differential analysis, and weighted gene co-expression network analysis (WGCNA) was applied to screen the differential co-expression genes associated with THCA tissue types. Univariant Cox regression analysis was further used for the selection of survival-related genes. Then, LASSO regression model was constructed to analyze the genes, and an optimal prognostic model was developed as well as evaluated by Kaplan-Meier and ROC curves. Results: Three thousand two hundred seven differentially expressed genes (DEGs) were obtained by differential analysis and 23 co-expression genes (|COR| > 0.5, P < 0.05) were gained after WGCNA analysis. In addition, eight genes significantly related to THCA survival were screened by univariant Cox regression analysis, and an optimal prognostic 3-gene risk prediction model was constructed after genes were analyzed by the LASSO regression model. Based on this model, patients were grouped into the high-risk group and low-risk group. Kaplan-Meier curve showed that patients in the low-risk group had much better survival than those in the high-risk group. Moreover, great accuracy of the 3-gene model was revealed by ROC curve and the remarkable correlation between the model and patients' prognosis was verified using the multivariant Cox regression analysis. Conclusion: The prognostic 3-gene model composed by GHR, GPR125, and ATP2C2 three genes can be used as an independent prognostic factor and has better prediction for the survival of THCA patients.