Lasso-Cox Model Research Articles

Pan-cancer evaluation of gene expression and somatic alteration data for cancer prognosis prediction

BackgroundOver the past decades, approaches for diagnosing and treating cancer have seen significant improvement. However, the variability of patient and tumor characteristics has limited progress on methods for prognosis prediction. The development of high-throughput omics technologies now provides multiple approaches for characterizing tumors. Although a large number of published studies have focused on integration of multi-omics data and use of pathway-level models for cancer prognosis prediction, there still exists a gap of knowledge regarding the prognostic landscape across multi-omics data for multiple cancer types using both gene-level and pathway-level predictors.MethodsIn this study, we systematically evaluated three often available types of omics data (gene expression, copy number variation and somatic point mutation) covering both DNA-level and RNA-level features. We evaluated the landscape of predictive performance of these three omics modalities for 33 cancer types in the TCGA using a Lasso or Group Lasso-penalized Cox model and either gene or pathway level predictors.ResultsWe constructed the prognostic landscape using three types of omics data for 33 cancer types on both the gene and pathway levels. Based on this landscape, we found that predictive performance is cancer type dependent and we also highlighted the cancer types and omics modalities that support the most accurate prognostic models. In general, models estimated on gene expression data provide the best predictive performance on either gene or pathway level and adding copy number variation or somatic point mutation data to gene expression data does not improve predictive performance, with some exceptional cohorts including low grade glioma and thyroid cancer. In general, pathway-level models have better interpretative performance, higher stability and smaller model size across multiple cancer types and omics data types relative to gene-level models.ConclusionsBased on this landscape and comprehensively comparison, models estimated on gene expression data provide the best predictive performance on either gene or pathway level. Pathway-level models have better interpretative performance, higher stability and smaller model size relative to gene-level models.

HIV transmission in sero-discordant couples with the HIV-positive partner in Xinjiang: incidence and associated predictors.

To control the spread of human immunodeficiency virus (HIV) among sero-discordant couples, we explored the HIV seroconversion and its contributing factors. We recruited negative partners in HIV sero-discordant couples to established a prospective cohort between January 2010 and June 2015 from areas with severe HIV epidemic in Xinjiang. Follow up once every 3 months, serological tests and risk behavior surveys every 6 months. Variables were screened by LASSO regression and a Cox proportional hazards model was established. A total of 1162 negative partners of sero-discordant couples were recruited. The seroconversion occurred in 42 negative partners during follow-up period, with a seroconversion rate of 2/100 (95% CI = 1.21-2.27), and the median time for seroconversion was 0.92 years. The Cox model showed that frequency of sexual behavior for nearly six months, consistent condom use, knowledge of the transmission route for HIV, a history of sexually transmitted diseases, recent CD4 + T lymphocyte count were all significant contributing factors to the seroconversion in negative partner of HIV sero-discordant couples. In addition, the Cox model was used to evaluate the risk factors of seroconversion for HIV negative partners. The seroconversion rate of HIV negative partners in Xinjiang was lower. The LASSO Cox model may accurately predict the risk of HIV transmission in sero-discordant couples.

Development of a tRNA-Derived Small RNA Prognostic Panel and Their Potential Functions in Osteosarcoma.

BackgroundTherapeutic outcomes of osteosarcoma treatment have not significantly improved in several decades. Therefore, strong prognostic biomarkers are urgently needed.MethodsWe first extracted the tRNA-derived small RNA (tsRNA) expression profiles of osteosarcoma from the GEO database. Then, we performed a unique module analysis and use the LASSO-Cox model to select survival-associated tsRNAs. Model effectiveness was further verified using an independent validation dataset. Target genes with selected tsRNAs were predicted using RNAhybrid.ResultsA LASSO-Cox model was established to select six prognostic tsRNA biomarkers: tRF-33-6SXMSL73VL4YDN, tRF-32-6SXMSL73VL4YK, tRF-32-M1M3WD8S746D2, tRF-35-RPM830MMUKLY5Z, tRF-33-K768WP9N1EWJDW, and tRF-32-MIF91SS2P46I3. We developed a prognostic panel for osteosarcoma patients concerning their overall survival by high-low risk. Patients with a low-risk profile had improved survival rates in training and validation dataset.ConclusionsThe suggested prognostic panel can be utilized as a reliable biomarker to predict osteosarcoma patient survival rates.

Dual-region radiomics signature: Integrating primary tumor and lymph node computed tomography features improves survival prediction in esophageal squamous cell cancer

BackgroundPreoperative prognostic biomarkers to guide individualized therapy are still in demand in esophageal squamous cell cancer (ESCC). Some studies reported that radiomic analysis based on CT images has been successfully performed to predict individual survival in EC. The aim of this study was to assess whether combining radiomics features from primary tumor and regional lymph nodes predicts overall survival (OS) better than using single-region features only, and to investigate the incremental value of the dual-region radiomics signature. MethodsIn this retrospective study, three radiomics signatures were built from preoperative enhanced CT in a training cohort (n = 200) using LASSO Cox model. Associations between each signature and survival was assessed on a validation cohort (n = 107). Prediction accuracy for the three signatures was compared. By constructing a clinical nomogram and a radiomics-clinical nomogram, incremental prognostic value of the radiomics signature over clinicopathological factors in OS prediction was assessed in terms of discrimination, calibration, reclassification and clinical usefulness. ResultsThe dual-region radiomic signature was an independent factor, significantly associated with OS (HR: 1.869, 95% CI: 1.347, 2.592, P = 1.82e-04), which achieved better OS (C-index: 0.611) prediction either than the single-region signature (C-index:0.594-0.604). The resulted dual-region radiomics-clinical nomogram achieved the best discriminative ability in OS prediction (C-index:0.700). Compared with the clinical nomogram, the radiomics-clinical nomogram improved the calibration and classification accuracy for OS prediction with a total net reclassification improvement (NRI) of 26.9% (P=0.008) and integrated discrimination improvement (IDI) of 6.8% (P<0.001). ConclusionThe dual-region radiomic signature is an independent prognostic marker and outperforms single-region signature in OS for ESCC patients. Integrating the dual-region radiomics signature and clinicopathological factors improves OS prediction.

Abstract 2149: Subgroups of early-stage microsatellite stability endometrial, colorectal and stomach tumors based on transcriptomics have distinct disease-free survival

Abstract Background: Microsatellite stability (MSS) is a biomarker for worse prognosis than microsatellite instability (MSI) in non-metastatic tumors. Tumors with MSS status are more probable to occur postoperative recurrence and metastasis. Some studies had revealed that endometrial and gastrointestinal tumors have distinct MSI-related features compared with other types of cancers, indicating there are potential similarities in these cancer types. The proportion of MSS population in non-metastatic endometrial and gastrointestinal tumors were about 90% and 75%. There are high heterogeneity in such wide range of patient population. The status of immune microenvironment based on immunohistochemistry or RNA-Seq profiling had been proved effective prognostic biomarker in some previous investigation. Therefore, we assume a subtype model of immune microenvironment based on transcriptomics features in endometrial and gastrointestinal tumors. Method: 128 patients with I-III stage MSS endometrial and gastrointestinal tumors from TCGA project were included as training cohort and a meta cohort of 603 patients of colorectal and stomach cancers from GEO datasets (GSE39582,GSE62254) were included as validation cohort in our study. The RNA-Seq profiling data and disease-free survival (DFS) of patients were collected. CIBERSORT tool was conducted to evaluate the immune-related cells' enrichment. A lasso-cox model of DFS was developed based on the CIBERSORT features. Patients in training cohort were scored the recurrence risk by the model. We further investigate the differentially expressed genes between high-risk and low-risk patients. Log-rank test was used to evaluate the difference of DFS. Result: 76 immune-related genes were identified to be differentially expressed significantly. The patients in training cohort were divided into two groups (group1 n=81, group2 n=41) based on non-negative matrix factorization (NMF) method. The DFS of group1 and group2 were significantly different (log-rank p=0.005). In validation cohort, patients were divided into group1 and group2 (group1 n=401, group2 n=202) using clustering method above. A significant difference between subgroups was also tested (log-rank p&amp;lt;0.0001). Conclusion: We identified 76 immune-related genes using CIBERSORT tool. Based on this gene set, early stage MSS endometrial and gastrointestinal tumors could be divided into two subgroups with distinct disease-free survival. Citation Format: Shi Zhang, Xinyi Liu, Lan Zhang, Mengli Huang. Subgroups of early-stage microsatellite stability endometrial, colorectal and stomach tumors based on transcriptomics have distinct disease-free survival [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 2149.

PCLasso: a protein complex-based, group lasso-Cox model for accurate prognosis and risk protein complex discovery.

For high-dimensional expression data, most prognostic models perform feature selection based on individual genes, which usually lead to unstable prognosis, and the identified risk genes are inherently insufficient in revealing complex molecular mechanisms. Since most genes carry out cellular functions by forming protein complexes-basic representatives of functional modules, identifying risk protein complexes may greatly improve our understanding of disease biology. Coupled with the fact that protein complexes have been shown to have innate resistance to batch effects and are effective predictors of disease phenotypes, constructing prognostic models and selecting features with protein complexes as the basic unit should improve the robustness and biological interpretability of the model. Here, we propose a protein complex-based, group lasso-Cox model (PCLasso) to predict patient prognosis and identify risk protein complexes. Experiments on three cancer types have proved that PCLasso has better prognostic performance than prognostic models based on individual genes. The resulting risk protein complexes not only contain individual risk genes but also incorporate close partners that synergize with them, which may promote the revealing of molecular mechanisms related to cancer progression from a comprehensive perspective. Furthermore, a pan-cancer prognostic analysis was performed to identify risk protein complexes of 19 cancer types, which may provide novel potential targets for cancer research.

Six-gene model based on expression predicts distinct disease-free survival in early-stage microsatellite stability (MSS) endometrial, colorectal and stomach tumors.

e14578 Background: Microsatellite stability (MSS) tumors hardly benefit from immunotherapies and are more probable to occur postoperative recurrence. However, some studies have revealed that a subset of MSS patients harbor “hot immune microenvironment” tumors, indicating high heterogeneity in such wide range of patient population. On the other hand, researches of mechanism of MSI formation found potential similarities in endometrial and gastrointestinal tumors. We hypothesized the transcriptomic features in these cancers correlated with immune-related signatures and patients’ prognosis. Methods: Early stage (I-III stage) MSS tumors, including endometrial, colorectal, and gastric from TCGA project were analyzed as training cohort (n=170). A combined cohort consisting of 604 colorectal and stomach cancers from GEO datasets (GSE39582，GSE62254) was validation cohort. The RNA-Seq profiling data and disease-free survival (DFS) data of patients were collected. Cibersort tool was used to evaluate twenty-two immune cells’ enrichment. The prediction model was developed by three steps: Univariate cox regression of DFS was conducted to select 9 immune cells. Then the train cohort was divided into two groups based on non-negative matrix factorization (NMF) method using this 9 immune cell features. Differentially expressed genes of these two groups were identified and screened further by lasso regression. Log-rank test was used to evaluate the difference of DFS. Results: A six-gene lasso-cox model was developed. The genes were LYZ, WFDC2, CAPS, RHPN1, TFF2 and TGFBR2. Based on the score evaluated by this model, patients in training cohort were divided into high-risk and low-risk groups. Low-risk population had much longer DFS (HR 0.07, 95%CI 0.03-0.18, p&lt;0.001). In validation cohort, lower risk score was also verified to be associated with a lower likelihood of recurrence (HR 0.66, 95%CI 0.5-0.88, p=0.0047). Conclusions: We developed a model of six-genes predicting disease-free survival based on RNA-Seq data in early stage MSS patients. Further validation was needed to implement in larger clinical cohorts.

Prognostic Models for Nonmetastatic Triple-Negative Breast Cancer Based on the Pretreatment Serum Tumor Markers with Machine Learning.

Purpose Triple-negative breast cancer (TNBC) is a heterogeneous and aggressive disease with poorer prognosis than other subtypes. We aimed to investigate the prognostic efficacy of multiple tumor markers and constructed a prognostic model for stage I-III TNBC patients. Patients and Methods. We included stage I-III TNBC patients whose serum tumor markers levels were measured prior to the treatment. The optimal cut-off value of each tumor marker was determined by X-tile. Then, we adopted two survival models (lasso Cox model and random survival forest model) to build the prognostic model and AUC values of the time-dependent receiver operating characteristic (ROC) were calculated. The Kaplan-Meier method was used to plot the survival curves and the log-rank test was used to test whether there was a significant difference between the predicted high-risk and low-risk groups. We used univariable and multivariable Cox analysis to identify independent prognostic factors and did subgroup analysis further for the lasso Cox model. Results We included 258 stage I-III TNBC patients. CEA, CA125, and CA211 showed independent prognostic value for DFS when using the optimal cut-off values; their HRs and 95% CI were as follows: 1.787 (1.056–3.226), 2.684 (1.200–3.931), and 2.513 (1.567–4.877). AUC values of lasso Cox model and random survival forest model were 0.740 and 0.663 for DFS at 60 months, respectively. Both the lasso Cox model and random survival forest model demonstrated excellent prognostic value. According to tumor marker risk scores (TMRS) computed by the lasso Cox model, the high TMRS group had worse DFS (HR = 3.138, 95% CI: 1.711–5.033, p < 0.0001) and OS (3.983, 1.637–7.214, p=0.0011) than low TMRS group. Furthermore, subgroup analysis of N0-N1 patients in the lasso Cox model indicated that TMRS still had a significant prognostic effect on DFS (2.278, 1.189–4.346) and OS (2.982, 1.110–7.519). Conclusions Our study indicated that pretreatment levels of serum CEA, CA125, and CA211 had independent prognostic significance for TNBC patients. Both lasso Cox model and random survival forest model that we constructed based on tumor markers could strongly predict the survival risk. Higher TMRS was associated with worse DFS and OS both in stage I-III and N0-N1 TNBC patients.

A fast adaptive Lasso for the cox regression via safe screening rules

Some interesting recent studies have shown that safe feature elimination screening algorithms are useful alternatives in solving large scale and/or ultra-high-dimensional Lasso-type problems. However, to the best of our knowledge, the plausibility of adapting the safe feature elimination screening algorithm to survival models is rarely explored. In this study, we first derive the safe feature elimination screening rule for adaptive Lasso Cox model. Then, using both simulated and real-world datasets, we demonstrate that the resulting algorithm can outperform Lasso Cox and adaptive Lasso Cox prediction methods in terms of its predictive performance. In addition to its good predictive performance, we illustrate that the proposed algorithm has a key computational advantage over the above competing methods in terms of computation efficiency.

Identifying Stage II Colorectal Cancer Recurrence Associated Genes by Microarray Meta-Analysis and Building Predictive Models with Machine Learning Algorithms.

Background Stage II colorectal cancer patients had heterogeneous prognosis, and patients with recurrent events had poor survival. In this study, we aimed to identify stage II colorectal cancer recurrence associated genes by microarray meta-analysis and build predictive models to stratify patients' recurrence-free survival. Methods We searched the GEO database to retrieve eligible microarray datasets. The microarray meta-analysis was used to identify universal recurrence associated genes. Total samples were randomly divided into the training set and the test set. Two survival models (lasso Cox model and random survival forest model) were trained in the training set, and AUC values of the time-dependent receiver operating characteristic (ROC) curves were calculated. Survival analysis was performed to determine whether there was significant difference between the predicted high and low risk groups in the test set. Results Six datasets containing 651 stage II colorectal cancer patients were included in this study. The microarray meta-analysis identified 479 recurrence associated genes. KEGG and GO enrichment analysis showed that G protein-coupled glutamate receptor binding and Hedgehog signaling were significantly enriched. AUC values of the lasso Cox model and the random survival forest model were 0.815 and 0.993 at 60 months, respectively. In addition, the random survival forest model demonstrated that the effects of gene expression on the recurrence-free survival probability were nonlinear. According to the risk scores computed by the random survival forest model, the high risk group had significantly higher recurrence risk than the low risk group (HR = 1.824, 95% CI: 1.079–3.084, p = 0.025). Conclusions We identified 479 stage II colorectal cancer recurrence associated genes by microarray meta-analysis. The random survival forest model which was based on the recurrence associated gene signature could strongly predict the recurrence risk of stage II colorectal cancer patients.

A Seven-Long Non-coding RNA Signature Improves Prognosis Prediction of Lung Adenocarcinoma: An Integrated Competing Endogenous RNA Network Analysis.

Early and precise prediction is an important way to reduce the poor prognosis of lung adenocarcinoma (LUAD) patients. Nevertheless, the widely used tumor, node, and metastasis (TNM) staging system based on anatomical information only often could not achieve adequate performance on foreseeing the prognosis of LUAD patients. This study thus aimed to examine whether the long non-coding RNAs (lncRNAs), known highly involved in the tumorigenesis of LUAD through the competing endogenous RNAs (ceRNAs) mechanism, could provide additional information to improve prognosis prediction of LUAD patients. To prove the hypothesis, a dataset consisting of both RNA sequencing data and clinical pathological data, obtained from The Cancer Genome Atlas (TCGA) database, was analyzed. Then, differentially expressed RNAs (DElncRNAs, DEmiRNAs, and DEmRNAs) were identified and a lncRNA–miRNA–mRNA ceRNA network was constructed based on those differentially expressed RNAs. Functional enrichment analysis revealed that this ceRNA network was highly enriched in some cancer-associated signaling pathways. Next, lasso-Cox model was run 1,000 times to recognize the potential survival-related combinations of the candidate lncRNAs in the ceRNA network, followed by the “best subset selection” to further optimize these lncRNA-based combinations, and a seven-lncRNA prognostic signature with the best performance was determined. Based on the median risk score, LUAD patients could be well distinguished into high-/low-risk subgroups. The Kaplan–Meier survival curve showed that LUAD patients in the high-risk group had significantly shorter overall survival than those in the low-risk group (log-rank test P = 4.52 × 10–9). The ROC curve indicated that the clinical genomic model including both the TNM staging system and the signature had a superior performance in predicting the patients’ overall survival compared to the clinical model with the TNM staging system only. Further stratification analysis suggested that the signature could work well in the different strata of the stage, gender, or age, rendering it to be a wide application. Finally, a ceRNA subnetwork related to the signature was extracted, demonstrating its high involvement in the tumorigenesis mechanism of LUAD. In conclusion, the present study established a lncRNA-based molecular signature, which can significantly improve prognosis prediction for LUAD patients.

Profiles of alternative splicing landscape in breast cancer and their clinical significance: an integrative analysis based on large-sequencing data

BackgroundAlternative splicing (AS) is closely correlated with the initiation and progression of carcinoma. The systematic analysis of its biological and clinical significance in breast cancer (BRCA) is, however, lacking.MethodsClinical data and RNA-seq were obtained from the TCGA dataset and differentially expressed AS (DEAS) events between tumor and paired normal BRCA tissues were identified. Enrichment analysis was then used to reveal the potential biological functions of DEAS events. We performed protein-protein interaction (PPI) analysis of DEAS events by using STRING and the correlation network between splicing factors (SFs) and AS events was constructed. The LASSO Cox model, Kaplan-Meier and log-rank tests were used to construct and evaluate DEAS-related risk signature, and the association between DEAS events and clinicopathological features were then analyzed.ResultsAfter strict filtering, 35,367 AS events and 973 DEAS events were detected. DEAS corresponding genes were significantly enriched in pivotal pathways including cell adhesion, cytoskeleton organization, and extracellular matrix organization. A total of 103 DEAS events were correlated with disease free survival. The DEAS-related risk signature stratified BRCA patients into two groups and the area under curve (AUC) was 0.754. Moreover, patients in the high-risk group had enriched basel-like subtype, advanced clinical stages, proliferation, and metastasis potency.ConclusionsCollectively, the profile of DEAS landscape in BRCA revealed the potential biological function and prognostic value of DEAS events.

A Panel of 6 Biomarkers Significantly Improves the Prediction of Type 2 Diabetes in the MONICA/KORA Study Population.

ContextImproved strategies to identify persons at high risk of type 2 diabetes are important to target costly preventive efforts to those who will benefit most.ObjectiveThis work aimed to assess whether novel biomarkers improve the prediction of type 2 diabetes beyond noninvasive standard clinical risk factors alone or in combination with glycated hemoglobin A1c (HbA1c).MethodsWe used a population-based case-cohort study for discovery (689 incident cases and 1850 noncases) and an independent cohort study (262 incident cases, 2549 noncases) for validation. An L1-penalized (lasso) Cox model was used to select the most predictive set among 47 serum biomarkers from multiple etiological pathways. All variables available from the noninvasive German Diabetes Risk Score (GDRSadapted) were forced into the models. The C index and the category-free net reclassification index (cfNRI) were used to evaluate the predictive performance of the selected biomarkers beyond the GDRSadapted model (plus HbA1c).ResultsInterleukin-1 receptor antagonist, insulin-like growth factor binding protein 2, soluble E-selectin, decorin, adiponectin, and high-density lipoprotein cholesterol were selected as the most relevant biomarkers. The simultaneous addition of these 6 biomarkers significantly improved the predictive performance both in the discovery (C index [95% CI], 0.053 [0.039-0.066]; cfNRI [95% CI], 67.4% [57.3%-79.5%]) and the validation study (0.034 [0.019-0.053]; 48.4% [35.6%-60.8%]). Significant improvements by these biomarkers were also seen on top of the GDRSadapted model plus HbA1c in both studies.ConclusionThe addition of 6 biomarkers significantly improved the prediction of type 2 diabetes when added to a noninvasive clinical model or to a clinical model plus HbA1c.

Integration of gene interaction information into a reweighted Lasso-Cox model for accurate survival prediction.

Accurately predicting the risk of cancer patients is a central challenge for clinical cancer research. For high-dimensional gene expression data, Cox proportional hazard model with the least absolute shrinkage and selection operator for variable selection (Lasso-Cox) is one of the most popular feature selection and risk prediction algorithms. However, the Lasso-Cox model treats all genes equally, ignoring the biological characteristics of the genes themselves. This often encounters the problem of poor prognostic performance on independent datasets. Here, we propose a Reweighted Lasso-Cox (RLasso-Cox) model to ameliorate this problem by integrating gene interaction information. It is based on the hypothesis that topologically important genes in the gene interaction network tend to have stable expression changes. We used random walk to evaluate the topological weight of genes, and then highlighted topologically important genes to improve the generalization ability of the RLasso-Cox model. Experiments on datasets of three cancer types showed that the RLasso-Cox model improves the prognostic accuracy and robustness compared with the Lasso-Cox model and several existing network-based methods. More importantly, the RLasso-Cox model has the advantage of identifying small gene sets with high prognostic performance on independent datasets, which may play an important role in identifying robust survival biomarkers for various cancer types. http://bioconductor.org/packages/devel/bioc/html/RLassoCox.html. Supplementary data are available at Bioinformatics online.

Identification of molecular mechanisms for achieving HIV-1 control in the absence of antiretroviral therapy

AimsAntiretroviral therapy (ART) controls viral replication but cannot eradicate an infected virus and restore the immune response of patients. Materials and methodsThe gene expression profiles of whole blood, PBMCs, CD4+ and CD8+ T cells were obtained from GSE108297. Coexpression analysis was carried out to evaluate differentially expressed genes (DEGs) between strong and weak responder HIV controllers (HICs). Enrichment analysis was used to explore the biological functions of DEGs. The key genes with common DEGs were screened using the Lasso Cox model. Then, the immune scores of HICs and HAART were calculated by ssGSEA. The content of CD4+ and CD8+ T cells, key genes were verified by flow cytometry, RT-PCR and Western blot analysis. Key findingsDEGs were clustered into 24 coexpression modules. DEGs related to general immune responses had the highest correlation with strong responding HICs, while DEGs mainly related to the apoptotic process had the highest correlation with weak responder HICs. The hub genes CD8A and CCT2, as well as the key genes TMEM132C and S100A9, were DEGs in HICs and HARRT. The immune score and flow cytometry showed that CD4+ and CD8+ T cells of HICs were lower than those of HARRT in whole blood. Experiments confirmed the expression of key genes in HICs and HARRT. SignificanceThe key genes identified in this study highlight the strong responder HICs features that to help the immune system control HIV-1 infection. These results will be useful for developing therapeutic targets.

Construction of a Comprehensive Diagnostic Scoring Model for Prostate Cancer Based on Six Gene-Panel

Accumulating evidence indicates that the N6-methyladenosine (m6A) modification plays a critical role in human cancers.Given the current understanding of m6A modification, this process is believed to be dynamically regulated by m6A regulators. Although the discovery of m6A regulators has greatly enhanced our understanding of the mechanism underlying m6A modification in cancers, the function and role of m6A in the context of prostate cancer (PCa) remains unclear.Here, we aimed to establish a comprehensive diagnostic scoring model that can be used to prostate-specific antigen (PSA) screening.To achieve this, we first drew the landscape of m6A regulators and constructed a LASSO-Cox model using 3 risk genes (METTL14, HNRNP2AB1 and YTHDF2). Particularly,METTL14 expression was found to be significantly related to overall survival, tumor T stage and relapse rate, showing that METTL14 has important prognostic value. Furthermore, for the sake of improving the predictive ability, we presented a comprehensive diagnostic scoring model based on 6 gene-panel by combining with genes found in our previous study.In conclusion, our study provides additional clues and insights regarding the treatment and diagnosis of PCa patients. Funding Statement: This work was supported by grants from the National Natural Science Foundation of China (31870928, Project-recepient: Q.L; 81770494, Project-recepient: Q. X.). Declaration of Interests: The authors declare that they have no conflicts of interest.

Identification of an Immune Signature Predicting Prognosis Risk and Lymphocyte Infiltration in Colon Cancer.

Increasing studies have highlighted the effects of the tumor immune micro-environment (TIM) on colon cancer (CC) tumorigenesis, prognosis, and metastasis. However, there is no reliable molecular marker that can effectively estimate the immune infiltration and predict the CC relapse risk. Here, we leveraged the gene expression profile and clinical characteristics from 1430 samples, including four gene expression omnibus database (GEO) databases and the cancer genome atlas (TCGA) database, to construct an immune risk signature that could be used as a predictor of survival outcome and immune activity. A risk model consisting of 10 immune-related genes were screened out in the Lasso-Cox model and were then aggregated to generate the immune risk signature based on the regression coefficients. The signature demonstrated robust prognostic ability in discovery and validation datasets, and this association remained significant in the multivariate analysis after controlling for age, gender, clinical stage, or microsatellite instability status. Leukocyte subpopulation analysis indicated that the low-risk signature was enriched with cytotoxic cells (activated CD4/CD8+ T cell and NK cell) and depleted of myeloid-derived suppressor cells (MDSC) and regulatory T cells. Further analysis indicated patients with a low-risk signature harbored higher tumor mutation loads and lower mutational frequencies in significantly mutated genes of APC and FBXW7. Together, our constructed signature could predict prognosis and represent the TIM of CC, which promotes individualized treatment and provides a promising novel molecular marker for immunotherapy.

LINC00649 underexpression is an adverse prognostic marker in acute myeloid leukemia

BackgroundLong noncoding RNAs (lncRNA) play a role in leukemogenesis, maintenance, development, and therapeutic resistance of AML. While few studies have focused on the prognostic significance of LINC00649 in AML, which we aim to investigate in this present study.MethodsWe compared the expression level of LINC00649 between AML patients and healthy controls. The Kaplan-Meier curves of AML patients expressing high versus low level of LINC00649 was performed. The LINC00649 correlated genes/miRNAs/lncRNAs and methylation CpG sites were screened by Pearson correlation analysis with R (version 3.6.0), using TCGA-LAML database. The LINC00649 associated ceRNA network was established using lncBase 2.0 and miRWalk 2.0 online tools, combining results from correlation analysis. Finally, a prediction model was constructed using LASSO-Cox regression.ResultsLINC00649 was underexpressed in bone marrow of AML group than that in healthy control group. The patients of LINC00649-low group have significantly inferior PFS and OS. A total of 154 mRNAs, 31 miRNAs, 28 lncRNAs and 1590 methylated CpG sites were identified to be significantly correlated with LINC00649. Furthermore, the network of ceRNA was established with 6 miRNAs and 122 mRNAs. The Lasso-Cox model fitted OS/PFS to novel prediction models, which integrated clinical factors, ELN risk stratification, mRNA/miRNA expression and methylation profiles. The analysis of time-dependent ROC for our model showed a superior AUC (AUC = 0.916 at 1 year, AUC = 0.916 at 3 years, and AUC = 0.891 at 5 years).ConclusionsLow expression of LINC00649 is a potential unfavorable prognostic marker for AML patients, which requires the further validation. The analysis by LASSO-COX regression identified a novel comprehensive model with a superior diagnostic utility, which integrated clinical and genetic variables.

Overexpression of HOXA10 is associated with unfavorable prognosis of acute myeloid leukemia

BackgroundHOXA family genes were crucial transcription factors involving cell proliferation and apoptosis. While few studies have focused on HOXA10 in AML. We aimed to investigate the prognostic significance of HOXA10.MethodsWe downloaded datasets from GEO and BeatAML database, to compare HOXA expression level between AML patients and controls. Kaplan-Meier curves were used to estimate the impact of HOXA10 expression on AML survival. The differentially expressed genes, miRNAs, lncRNAs and methylated regions between HOXA10-high and -low groups were obtained using R (version 3.6.0). Accordingly, the gene set enrichment analysis (GSEA) was accomplished using MSigDB database. Moreover, the regulatory TFs/microRNAs/lncRNAs of HOXA10 were identified. A LASSO-Cox model fitted OS to clinical and HOXA10-associated genetic variables by glmnet package.ResultsHOXA10 was overexpressed in AML patients than that in controls. The HOXA10-high group is significantly associated with shorter OS and DFS. A total of 1219 DEGs, 131 DEmiRs, 282 DElncRs were identified to be associated with HOXA10. GSEA revealed that 12 suppressed and 3 activated pathways in HOXA10-high group. Furthermore, the integrated regulatory network targeting HOXA10 was established. The LASSO-Cox model fitted OS to AML-survival risk scores, which included age, race, molecular risk, expression of IKZF2/LINC00649/LINC00839/FENDRR and has-miR-424-5p. The time dependent ROC indicated a satisfying AUC (1-year AUC 0.839, 3-year AUC 0.871 and 5-year AUC 0.813).ConclusionsOur study identified HOXA10 overexpression as an adverse prognostic factor for AML. The LASSO-COX regression analysis revealed novel prediction model of OS with superior diagnostic utility.

Survival prediction model for non-small cell lung cancer based on somatic mutations.

The 5-year survival rate of non-small cell lung cancer (NSCLC) is only 15%. Screening some combined gene mutations could predict the survival of NSCLC patients and also provide new ideas for the diagnosis and treatment of NSCLC. The present study aimed to identify signature mutations for survival prediction of NSCLC. Clinical and gene mutation information for 949 NSCLC patients was downloaded from TCGA. High frequency mutation and common mutation genes were analyzed based on 1000 cancer related genes. The LASSO-COX model was used to screen gene mutation points and analyze their survival, and then a survival prediction model was established. Fifty NSCLC patients were collected and 1000 targeted genes were enriched by targeted next generation sequencing. The results were used to verify the combination of common mutation genes and the function of the survival model, and then to clarify their clinical significance. Ten variables were screened out after LASSO-COX analysis, including age, tumor stage, EGFR c.[2,573 T>G], PIK3CA c.[1624G>A], TP53 c.[375G>T], TP53 c.[527G>T], TP53 c.[733G>T], TP53 c.[734G>T], TP53 c.[743G>T], NFE2L2 c.[100C>G]. Except for TP53 c.[743G>T] and NFE2L2 c.[100C>G], the residual six hot spot mutations of EGFR, PIK3CA and TP53 could be regarded as a signature mutations for forecasting the survival time of NSCLC. The combination of six hot spot mutations of EGFR, PIK3CA and TP53 is expected to be used for predicting the survival time of NSCLC.