Prognostic model and immune-infiltrating cell landscape based on differentially expressed autophagy-related genes in TP53-mutated multiple myeloma

IntroductionAutophagy functions as a prosurvival mechanism in multiple myeloma (MM).The objective of this research was to establish an autophagy-related gene (ARG) signature for predicting the survival outcomes of MM patients with TP53 mutations.Material and methodsInformation about MM patients with TP53 mutations was downloaded from Gene Expression Omnibus (GEO) database. Cox proportional hazard regression was employed to determine the independent prognostic ARG and construct a risk signature. Time-dependent receiver-operating characteristic (t ROC) curve was used to explore the predictive accuracy of the prognostic model. A nomogram was constructed to give a more precise prediction of the probability of 5-year, 8-year and 10-year overall survival (OS). In addition, we utilized the CIBERSORT algorithm to explore the distribution difference of 22 immune-infiltrating cells.ResultsThree differentially expressed ARGs (CASP8, MAPK8, RB1CC1) were finally incorporated to construct the risk model. Area under the curve (AUC) of corresponding tROC curve for 5-year,8-year and 10-year OS were 0.735, 0.686 and 0.662, respectively. MM patients were categorized into high and low-risk group in accordance with the median threshold value (-1.724549). ARG-based risk score model was an independent prognostic element correlated with OS, giving an hazard ratio (HR) of 3.29 (95%CI 2.35-4.60, P<0.001). 13 immune infiltrating cells were found to have distribution differences between the two groups.ConclusionsWe established a three-ARGs risk signature which manifested an independent prognostic factor. The nomogram was testified to perform well in forecasting the long-term survival of TP53-mutated MM patients.

Download Full-text

Prognostic Value of Plasma hPG80 (Circulating Progastrin) in Metastatic Renal Cell Carcinoma

Cancers ◽

10.3390/cancers13030375 ◽

2021 ◽

Vol 13 (3) ◽

pp. 375

Author(s):

Manish Kohli ◽

Winston Tan ◽

Bérengère Vire ◽

Pierre Liaud ◽

Mélina Blairvacq ◽

...

Keyword(s):

Renal Cell Carcinoma ◽

Cell Carcinoma ◽

Renal Cell ◽

Prognostic Value ◽

Risk Model ◽

Predictive Accuracy ◽

Area Under The Curve ◽

Risk Groups ◽

Prognostic Impact ◽

Significant Difference

Precise management of kidney cancer requires the identification of prognostic factors. hPG80 (circulating progastrin) is a tumor promoting peptide present in the blood of patients with various cancers, including renal cell carcinoma (RCC). In this study, we evaluated the prognostic value of plasma hPG80 in 143 prospectively collected patients with metastatic RCC (mRCC). The prognostic impact of hPG80 levels on overall survival (OS) in mRCC patients after controlling for hPG80 levels in non-cancer age matched controls was determined and compared to the International Metastatic Database Consortium (IMDC) risk model (good, intermediate, poor). ROC curves were used to evaluate the diagnostic accuracy of hPG80 using the area under the curve (AUC). Our results showed that plasma hPG80 was detected in 94% of mRCC patients. hPG80 levels displayed high predictive accuracy with an AUC of 0.93 and 0.84 when compared to 18–25 year old controls and 50–80 year old controls, respectively. mRCC patients with high hPG80 levels (>4.5 pM) had significantly lower OS compared to patients with low hPG80 levels (<4.5 pM) (12 versus 31.2 months, respectively; p = 0.0031). Adding hPG80 levels (score of 1 for patients having hPG80 levels > 4.5 pM) to the six variables of the IMDC risk model showed a greater and significant difference in OS between the newly defined good-, intermediate- and poor-risk groups (p = 0.0003 compared to p = 0.0076). Finally, when patients with IMDC intermediate-risk group were further divided into two groups based on hPG80 levels within these subgroups, increased OS were observed in patients with low hPG80 levels (<4.5 pM). In conclusion, our data suggest that hPG80 could be used for prognosticating survival in mRCC alone or integrated to the IMDC score (by adding a variable to the IMDC score or by substratifying the IMDC risk groups), be a prognostic biomarker in mRCC patients.

Download Full-text

Identification of a three-gene-based prognostic model in multiple myeloma using bioinformatics analysis

PeerJ ◽

10.7717/peerj.11320 ◽

2021 ◽

Vol 9 ◽

pp. e11320

Author(s):

Ying Pan ◽

Ye Meng ◽

Zhimin Zhai ◽

Shudao Xiong

Keyword(s):

Gene Expression ◽

Multiple Myeloma ◽

Normal Control ◽

Prognostic Model ◽

Risk Model ◽

Cox Regression ◽

Expression Profiles ◽

Clinical Parameters ◽

Hub Genes ◽

Key Genes

Background Multiple myeloma (MM), the second most hematological malignancy, has high incidence and remains incurable till now. The pathogenesis of MM is poorly understood. This study aimed to identify novel prognostic model for MM on gene expression profiles. Methods Gene expression datas of MM (GSE6477, GSE136337) were downloaded from Gene Expression Omnibus (GEO) database. The differentially expressed genes (DEGs) in GSE6477 between case samples and normal control samples were screened by the limma package. Meanwhile, enrichment analysis was conducted, and a protein-protein interaction (PPI) network of these DEGs was established by STRING and cytoscape software. Co-expression modules of genes were built by Weighted Correlation Network Analysis (WGCNA). Key genes were identified both from hub genes and the DEGs. Univariate and multivariate Cox congression were performed to screen independent prognostic genes to construct a predictive model. The predictive power of the model was evaluated by Kaplan–Meier curve and time-dependent receiver operating characteristic (ROC) curves. Finally, univariate and multivariate Cox regression analyse were used to investigate whether the prognostic model could be independent of other clinical parameters. Results GSE6477, including 101 case and 15 normal control, were screened as the datasets. A total of 178 DEGs were identified, including 59 up-regulated and 119 down-regulated genes. In WGCNA analysis, module black and module purple were the most relevant modules with cancer traits, and 92 hub genes in these two modules were selected for further analysis. Next, 47 genes were chosen both from the DEGs and hub genes as key genes. Three genes (LYVE1, RNASE1, and RNASE2) were finally screened by univariate and multivariate Cox regression analyses and used to construct a risk model. In addition, the three-gene prognostic model revealed independent and accurate prognostic capacity in relation to other clinical parameters for MM patients. Conclusion In summary, we identified and constructed a three-gene-based prognostic model that could be used to predict overall survival of MM patients.

Download Full-text

Integrated analysis of the functions and prognostic values of RNA-binding proteins in neuroblastoma

PLoS ONE ◽

10.1371/journal.pone.0260876 ◽

2021 ◽

Vol 16 (12) ◽

pp. e0260876

Author(s):

Jun Yang ◽

Jiaying Zhou ◽

Cuili Li ◽

Shaohua Wang

Keyword(s):

Regression Analysis ◽

Risk Score ◽

Prognostic Model ◽

Binding Proteins ◽

Rna Binding ◽

Risk Model ◽

Cox Regression ◽

Rna Binding Proteins ◽

Differentially Expressed ◽

Lasso Regression

Background Neuroblastoma (NB) is the most common solid tumor in children. NB treatment has made significant progress; however, given the high degree of heterogeneity, basic research findings and their clinical application to NB still face challenges. Herein, we identify novel prognostic models for NB. Methods We obtained RNA expression data of NB and normal nervous tissue from TARGET and GTEx databases and determined the differential expression patterns of RNA binding protein (RBP) genes between normal and cancerous tissues. Lasso regression and Cox regression analyses identified the five most important differentially expressed genes and were used to construct a new prognostic model. The function and prognostic value of these RBPs were systematically studied and the predictive accuracy verified in an independent dataset. Results In total, 348 differentially expressed RBPs were identified. Of these, 166 were up-regulated and 182 down-regulated RBPs. Two hubs RBPs (CPEB3 and CTU1) were identified as prognostic-related genes and were chosen to build the prognostic risk score models. Multivariate Cox analysis was performed on genes from univariate Cox regression and Lasso regression analysis using proportional hazards regression model. A five gene prognostic model: Risk score = (-0.60901*expCPEB3)+(0.851637*expCTU1) was built. Based on this model, the overall survival of patients in the high-risk subgroup was lower (P = 2.152e-04). The area under the curve (AUC) of the receiver-operator characteristic curve of the prognostic model was 0.720 in the TARGET cohort. There were significant differences in the survival rate of patients in the high and low-risk subgroups in the validation data set GSE85047 (P = 0.1237e-08), with the AUC 0.730. The risk model was also regarded as an independent predictor of prognosis (HR = 1.535, 95% CI = 1.368–1.722, P = 2.69E-13). Conclusions This study identified a potential risk model for prognosis in NB using Cox regression analysis. RNA binding proteins (CPEB3 and CTU1) can be used as molecular markers of NB.

Download Full-text

Development, Prediction, and Treatment of Anemia in Patients (pts) with Lymphoma/Multiple Myeloma (L/M): Findings of Two European Surveys (ECAS and BEPOS).

Blood ◽

10.1182/blood.v104.11.3133.3133 ◽

2004 ◽

Vol 104 (11) ◽

pp. 3133-3133

Author(s):

Heinz Ludwig ◽

Simon Van Belle ◽

Pere Gascón

Keyword(s):

At Risk ◽

Multiple Myeloma ◽

Treatment Initiation ◽

Risk Model ◽

Predictive Accuracy ◽

Hematologic Malignancy ◽

Performance Status ◽

Female Gender ◽

Mean Time ◽

Anemia Treatment

Abstract Quality of life (QOL) in cancer pts is adversely affected by several factors, including disease- or treatment-related anemia. Between January 2001 and February 2002, the European Cancer Anaemia Survey (ECAS) was conducted to provide information on pervasiveness of cancer-related anemia, impact of anemia on WHO performance status, risk factors for its development, and anemia treatment practices in Europe. Briefly, a total of 15,370 pts were enrolled, of whom 2,360 had L/M. Data analysis showed 53% of L/M pts were anemic (hemoglobin [Hb] <12 g/dL) at enrollment and 73% were anemic at some time during the survey. Low Hb levels correlated significantly with WHO performance scores of 3 or 4 (P <0.001, R =.352). The ECAS data were additionally analyzed to determine patient and disease characteristics that predicted anemia development and to construct a model for identifying pts at risk. Using logistic regression on the L/M incidence group (pts who were not anemic and not being treated for cancer at enrollment, started chemotherapy [CT] during ECAS, and had at least 2 CT cycles during the survey), 4 variables were found to significantly predict anemia development. Initial Hb, persistent/recurrent disease, female gender, and intent to treat or treatment with platinum-based CT were found to independently predict anemia (P <0.001), with an area under the ROC curve of 0.821 (95% CI; 0.763–0.878), indicating acceptable predictive accuracy of the model. Three levels of risk (low [24%], moderate [51%], and high [72%]) for developing anemia were calculated from the model (χ2(2) = 112.6, P <0.001). Mean time required for anemia development was 9.0 wks to reach Hb of <12 g/dL, 11.1 wks to reach Hb of <11 g/dL, and 13.3 wks to reach Hb of <10 g/dL. Notably, only 46% of anemic L/M pts received anemia treatment. Subsequently, a recent (2003) survey, the Belgian Erythropoietin Survey (BEPOS), extended the information gained through ECAS by examining use of recombinant human erythropoietin (rHuEPO) in pts receiving CT. Specifically, BEPOS documented when rHuEPO treatment was started, dosing schedules and dosage adjustments, length of treatment, impact of iron supplementation on rHuEPO treatment, and outcomes. Patients enrolled had either solid tumors (non-small-cell lung cancer, breast cancer) or hematologic malignancies (multiple myeloma [MM], Hodgkin’s disease [HD], non-Hodgkin’s lymphoma [NHL]). Interim results suggest that 72% of BEPOS pts began rHuEPO during the first 2 CT cycles, with an overall median Hb value of 10.1 g/dL at treatment initiation. For hematologic malignancy pts, the median Hb at treatment initiation was ~10 g/dL for HD; in NHL and MM, the median Hb was >9 to <10 g/dL. Mean time for all pts to achieve a 2-g/dL increase in Hb in the absence of transfusion was 6.4 wks; mean time for NHL pts (6.3 wks) was similar, while mean time for MM pts (9.1 wks) was longer, more in line with that seen in clinical trials. Achievement of the 2-g/dL increase in Hb after 6.4 wks determined in BEPOS is consistent with increases of ~1 g/dL after 4 wks and ~2 g/dL after 8 wks noted in studies of epoetin alfa (Demetri 1998, Gabrilove 2001, Littlewood 2001). Using the large ECAS database, an anemia risk model has been established that should help identify pts at risk for anemia, so that administration of rHuEPO can be initiated expeditiously, before Hb declines to considerably lower levels and/or anemia symptoms, including impaired QOL, develop.

Download Full-text

Addition of eGFR and Age Improves the Prognostic Absolute Renal Risk-Model in 1,134 Norwegian Patients with IgA Nephropathy

American Journal of Nephrology ◽

10.1159/000381403 ◽

2015 ◽

Vol 41 (3) ◽

pp. 210-219 ◽

Cited By ~ 15

Author(s):

Thomas Knoop ◽

Ann Merethe Vågane ◽

Bjørn Egil Vikse ◽

Einar Svarstad ◽

Bergrún Tinna Magnúsdóttir ◽

...

Keyword(s):

Iga Nephropathy ◽

Prognostic Model ◽

Risk Model ◽

Area Under The Curve ◽

Predicting Outcome ◽

The Mean ◽

Stage Renal Disease ◽

End Stage ◽

French Cohort

Background: Predicting outcome in individual patients with IgA nephropathy (IgAN) is difficult but important. For this purpose, the absolute renal risk (ARR) model has been developed in a French cohort to calculate the risk of end-stage renal disease (ESRD) and death. ARR (0-3) is scored in individual IgAN patients based on the presence of proteinuria ≥1 g/24 h, hypertension, and severe histopathological lesions (1 point per risk factor). We have validated the ARR model in a Norwegian cohort of IgAN patients and tested whether adding data on initial estimated glomerular filtration rate (eGFR) and age improved prediction. Methods: IgAN patients diagnosed between 1988 and 2012 were identified in the Norwegian Kidney Biopsy Registry, and endpoints were identified by record linkage with the Norwegian Renal Registry (ESRD) and the Population Registry (deaths). Results: We identified 1,134 IgAN patients. The mean duration of follow-up was 10.2 years (range 0.0 to 25.7 years). Two hundred and fifty one patients developed ESRD and there were 69 pre-ESRD deaths. The ARR model significantly stratified the IgAN cohort according to risk of ESRD/death. The inclusion of eGFR and age significantly improved the ARR prognostic model; in the receiver operator characteristics (ROC) analysis, area under the curve (AUC) at 10-years of follow-up increased from 0.79 to 0.89, p < 0.001. Conclusions: ARR is a suitable prognostic model for stratifying IgAN patients according to the risk of ESRD or death. Including initial eGFR and age in the model substantially improved its accuracy in our nationwide cohort.

Download Full-text

A prognostic model for hepatocellular carcinoma based on apoptosis-related genes

10.21203/rs.3.rs-141169/v2 ◽

2021 ◽

Author(s):

Renjie Liu ◽

Guifu Wang ◽

Chi Zhang ◽

Dousheng Bai

Keyword(s):

Hepatocellular Carcinoma ◽

Prognostic Model ◽

Molecular Mechanisms ◽

Risk Model ◽

Cox Regression ◽

Risk Groups ◽

Low Risk ◽

The Cancer Genome Atlas ◽

Differentially Expressed ◽

Risk Scores

Abstract Background: Dysregulation of the balance between proliferation and apoptosis is the basis for human hepatocarcinogenesis. In many malignant tumors, such as hepatocellular carcinoma (HCC), there is a correlation between apoptotic dysregulation and poor prognosis. However, the prognostic values of apoptosis-related genes (ARGs) in HCC have not been elucidated. Methods: To screen for differentially expressed ARGs, the expression levels of 161 ARGs from The Cancer Genome Atlas (TCGA) database(https://cancergenome.nih.gov/) were analyzed. Gene Ontology (GO) enrichment and the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analyses were performed to evaluate the underlying molecular mechanisms of differentially expressed ARGs in HCC. The prognostic values of ARGs were established using Cox regression, and subsequently, a prognostic risk model for scoring patients was developed. Kaplan-Meier (K-M) and receiver operating characteristic (ROC) curves were plotted to determine the prognostic value of the model. Results: Compared to normal tissues, 43 highly up-regulated and 8 down-regulated ARGs in HCC tissues were screened. GO analysis results revealed that these 51 genes are indeed related to the apoptosis function. KEGG analysis revealed that these 51 genes were correlated with MAPK, P53, TNF, and PI3K-AKT signaling pathways, while Cox regression revealed that 5 ARGs (PPP2R5B, SQSTM1, TOP2A, BMF, and LGALS3) were associated with prognosis and were, therefore, obtained to develop the prognostic model. Based on the median risk scores, patients were categorized into high-risk and low-risk groups. Patients in the low-risk groups exhibited significantly elevated two-year or five-year survival probabilities (p < 0.0001). The risk model had a better clinical potency than the other clinical characteristics, with the area under the ROC curve (AUC = 0.741). The prognosis of HCC patients was established from a plotted nomogram. Conclusion: Based on the differential expression of ARGs, we established a novel risk model for predicting HCC prognosis. This model can also be used to inform the individualized treatment of HCC patients.

Download Full-text

Adjusting preoperative risk models of post heart transplant survival to a European cohort in the age of a new cardiac allocation score in Europe

The Heart Surgery Forum ◽

10.1532/hsf.2205 ◽

2018 ◽

Vol 21 (6) ◽

pp. E527-E533

Author(s):

Tarik Alp Sargut ◽

Panagiotis Pergantis ◽

Christoph Knosalla ◽

Jan Knierim ◽

Manfred Hummel ◽

...

Keyword(s):

Heart Transplant ◽

Scoring System ◽

Graft Failure ◽

Risk Model ◽

Predictive Accuracy ◽

Area Under The Curve ◽

Risk Scores ◽

Risk Models ◽

Adjusted Risk ◽

The Impact

Background Several risk models target the issue of posttransplant survival, but none of them have been validated in a large European cohort. This aspect is important, in a time of the planned change of the Eurotransplant allocation system to a scoring system. Material and Methods Data of 761 heart transplant recipients from the Eurotransplant region with a total follow up of 5027 patient-years were analyzed. We assessed 30-day to 10-year freedom from graft failure. Existing post-transplant mortality risk models, IMPACT, Meld-XI and Columbia Risk Stratification Score were (RSS) were evaluated. A new risk model was created and the predictive accuracy was compared with the existing risk scores, with a focus on LVAD patients. Results Thirty-day, 1-year, 5-year and 10-year rates of freedom from graft failure were 78.3±1.5%, 68.8±1.71%, 59.1±1.8% and 44.1±1.9. The 1-year incidence of graft failure varied from 14.1% to 50% (RSS), from 22.9% to 57.1 (IMPACT) and from 24.9% to 42.6% using MELD-XI. Our newly adjusted risk score showed an improved area under the curve (AUC) of 0.69 (95% CI 0.64-0.72) with better discrimination in the intermediate to moderate risk cohort (CABDES Score). Conclusion IMPACT, Meld-XI and RSS were suitable to predict posttransplant graft failure only in a high and low risk cohort. CABDES Score, might be an alternative scoring system, with donor age and eGFR beeing the strongest predictors. Implementation of the IMPACT score within the new Eurotransplant Cardiac Allocation Score for patient prioritization for heart transplantation, should be reevaluated.

Download Full-text

A prognostic model for hepatocellular carcinoma based on apoptosis-related genes

World Journal of Surgical Oncology ◽

10.1186/s12957-021-02175-9 ◽

2021 ◽

Vol 19 (1) ◽

Author(s):

Renjie Liu ◽

Guifu Wang ◽

Chi Zhang ◽

Dousheng Bai

Keyword(s):

Hepatocellular Carcinoma ◽

Prognostic Model ◽

Molecular Mechanisms ◽

Risk Model ◽

Cox Regression ◽

Risk Groups ◽

Low Risk ◽

The Cancer Genome Atlas ◽

Differentially Expressed ◽

Risk Scores

Abstract Background Dysregulation of the balance between proliferation and apoptosis is the basis for human hepatocarcinogenesis. In many malignant tumors, such as hepatocellular carcinoma (HCC), there is a correlation between apoptotic dysregulation and poor prognosis. However, the prognostic values of apoptosis-related genes (ARGs) in HCC have not been elucidated. Methods To screen for differentially expressed ARGs, the expression levels of 161 ARGs from The Cancer Genome Atlas (TCGA) database (https://cancergenome.nih.gov/) were analyzed. Gene Ontology (GO) enrichment and the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analyses were performed to evaluate the underlying molecular mechanisms of differentially expressed ARGs in HCC. The prognostic values of ARGs were established using Cox regression, and subsequently, a prognostic risk model for scoring patients was developed. Kaplan–Meier (K-M) and receiver operating characteristic (ROC) curves were plotted to determine the prognostic value of the model. Results Compared with normal tissues, 43 highly upregulated and 8 downregulated ARGs in HCC tissues were screened. GO analysis results revealed that these 51 genes are indeed related to the apoptosis function. KEGG analysis revealed that these 51 genes were correlated with MAPK, P53, TNF, and PI3K-AKT signaling pathways, while Cox regression revealed that 5 ARGs (PPP2R5B, SQSTM1, TOP2A, BMF, and LGALS3) were associated with prognosis and were, therefore, obtained to develop the prognostic model. Based on the median risk scores, patients were categorized into high-risk and low-risk groups. Patients in the low-risk groups exhibited significantly elevated 2-year or 5-year survival probabilities (p < 0.0001). The risk model had a better clinical potency than the other clinical characteristics, with the area under the ROC curve (AUC = 0.741). The prognosis of HCC patients was established from a plotted nomogram. Conclusion Based on the differential expression of ARGs, we established a novel risk model for predicting HCC prognosis. This model can also be used to inform the individualized treatment of HCC patients.

Download Full-text

Validation and verification of predictive salivary biomarkers for oral health

Scientific Reports ◽

10.1038/s41598-021-85120-w ◽

2021 ◽

Vol 11 (1) ◽

Cited By ~ 1

Author(s):

Nagihan Bostanci ◽

Konstantinos Mitsakakis ◽

Beral Afacan ◽

Kai Bao ◽

Benita Johannsen ◽

...

Keyword(s):

Oral Health ◽

Predictive Accuracy ◽

Statistical Tests ◽

Area Under The Curve ◽

Regression Tree ◽

High Sensitivity ◽

Oral Health Status ◽

Classification And Regression Tree ◽

Cart Analysis ◽

Salivary Biomarkers

AbstractOral health is important not only due to the diseases emerging in the oral cavity but also due to the direct relation to systemic health. Thus, early and accurate characterization of the oral health status is of utmost importance. There are several salivary biomarkers as candidates for gingivitis and periodontitis, which are major oral health threats, affecting the gums. These need to be verified and validated for their potential use as differentiators of health, gingivitis and periodontitis status, before they are translated to chair-side for diagnostics and personalized monitoring. We aimed to measure 10 candidates using high sensitivity ELISAs in a well-controlled cohort of 127 individuals from three groups: periodontitis (60), gingivitis (31) and healthy (36). The statistical approaches included univariate statistical tests, receiver operating characteristic curves (ROC) with the corresponding Area Under the Curve (AUC) and Classification and Regression Tree (CART) analysis. The main outcomes were that the combination of multiple biomarker assays, rather than the use of single ones, can offer a predictive accuracy of > 90% for gingivitis versus health groups; and 100% for periodontitis versus health and periodontitis versus gingivitis groups. Furthermore, ratios of biomarkers MMP-8, MMP-9 and TIMP-1 were also proven to be powerful differentiating values compared to the single biomarkers.

Download Full-text

Mining the plasma-proteome associated genes in patients with gastro-esophageal cancers for biomarker discovery

Scientific Reports ◽

10.1038/s41598-021-87037-w ◽

2021 ◽

Vol 11 (1) ◽

Author(s):

Frederick S. Vizeacoumar ◽

Hongyu Guo ◽

Lynn Dwernychuk ◽

Adnan Zaidi ◽

Andrew Freywald ◽

...

Keyword(s):

Biomarker Discovery ◽

Predictive Accuracy ◽

Area Under The Curve ◽

Cancer Recurrence ◽

The Cancer Genome Atlas ◽

Svm Classifier ◽

Plasma Proteome ◽

Cancer Genome Atlas ◽

Novel Biomarkers ◽

Evaluation Metric

AbstractGastro-esophageal (GE) cancers are one of the major causes of cancer-related death in the world. There is a need for novel biomarkers in the management of GE cancers, to yield predictive response to the available therapies. Our study aims to identify leading genes that are differentially regulated in patients with these cancers. We explored the expression data for those genes whose protein products can be detected in the plasma using the Cancer Genome Atlas to identify leading genes that are differentially regulated in patients with GE cancers. Our work predicted several candidates as potential biomarkers for distinct stages of GE cancers, including previously identified CST1, INHBA, STMN1, whose expression correlated with cancer recurrence, or resistance to adjuvant therapies or surgery. To define the predictive accuracy of these genes as possible biomarkers, we constructed a co-expression network and performed complex network analysis to measure the importance of the genes in terms of a ratio of closeness centrality (RCC). Furthermore, to measure the significance of these differentially regulated genes, we constructed an SVM classifier using machine learning approach and verified these genes by using receiver operator characteristic (ROC) curve as an evaluation metric. The area under the curve measure was > 0.9 for both the overexpressed and downregulated genes suggesting the potential use and reliability of these candidates as biomarkers. In summary, we identified leading differentially expressed genes in GE cancers that can be detected in the plasma proteome. These genes have potential to become diagnostic and therapeutic biomarkers for early detection of cancer, recurrence following surgery and for development of targeted treatment.

Download Full-text