RÉSUMÉ
In the realm of colon carcinoma, significant genetic and epigenetic diversity is observed, underscoring the necessity for tailored prognostic features that can guide personalized therapeutic strategies. In this study, we explored the association between the type 2 bitter taste receptor (TAS2Rs) family-related genes and colon cancer using RNA-sequencing and clinical datasets from The Cancer Genome Atlas (TCGA) and the Gene Expression Omnibus (GEO). Our preliminary analysis identified seven TAS2Rs genes associated with survival using univariate Cox regression analysis, all of which were observed to be overexpressed in colon cancer. Subsequently, based on these seven TAS2Rs prognostic genes, two colon cancer molecular subtypes (Cluster A and Cluster B) were defined. These subtypes exhibited distinct prognostic and immune characteristics, with Cluster A characterized by low immune cell infiltration and less favorable outcomes, while Cluster B was associated with high immune cell infiltration and better prognosis. Finally, we developed a robust scoring system using a gradient boosting machine (GBM) approach, integrated with the gene-pairing method, to predict the prognosis of colon cancer patients. This machine learning model could improve our predictive accuracy for colon cancer outcomes, underscoring its value in the precision oncology framework.
Sujet(s)
Tumeurs du côlon , Régulation de l'expression des gènes tumoraux , Récepteurs couplés aux protéines G , Humains , Tumeurs du côlon/génétique , Tumeurs du côlon/mortalité , Tumeurs du côlon/anatomopathologie , Pronostic , Récepteurs couplés aux protéines G/génétique , Marqueurs biologiques tumoraux/génétique , Femelle , Apprentissage machine , Analyse de profil d'expression de gènes , MâleRÉSUMÉ
Background: Colon cancer is a highly heterogeneous disease, and identifying molecular subtypes can provide insights into deregulated pathways within tumor subsets, which may lead to personalized treatment options. However, most prognostic models are based on single-pathway genes. Methods: In this study, we aimed to identify three clinically relevant subtypes of colon cancer based on multiple signaling pathways-related genes. Integrative multi-omics analysis was used to explain the biological processes contributing to colon cancer aggressiveness, recurrence, and progression. Machine learning methods were employed to identify the subtypes and provide medication guidance for distinct subtypes using the L1000 platform. We developed a robust prognostic model (MKPC score) based on gene pairs and validated it in one internal test set and three external test sets. Risk-related genes were extracted and verified by qPCR. Results: Three clinically relevant subtypes of colon cancer were identified based on multiple signaling pathways-related genes, which had significantly different survival state (Log-Rank test, p<0.05). Integrative multi-omics analysis revealed biological processes contributing to colon cancer aggressiveness, recurrence, and progression. The developed MKPC score, based on gene pairs, was robust in predicting prognosis state (Log-Rank test, p<0.05), and risk-related genes were successfully verified by qPCR (t test, p<0.05). An easy-to-use web tool was created for risk scoring and therapy stratification in colon cancer patients, and the practical nomogram can be extended to other cancer types. Conclusion: In conclusion, our study identified three clinically relevant subtypes of colon cancer and developed a robust prognostic model based on gene pairs. The developed web tool is a valuable resource for researchers and clinicians in risk scoring and therapy stratification in colon cancer patients, and the practical nomogram can be extended to other cancer types.
Sujet(s)
Tumeurs du côlon , Multi-omique , Humains , Immunothérapie , Analyse de données , Apprentissage machineRÉSUMÉ
We used an integrated ensemble learning method to build a stable prediction model for severity in COVID-19 patients, which was validated in multicenter cohorts.
RÉSUMÉ
The molecular landscape in breast cancer is characterized by large biological heterogeneity and variable clinical outcomes. Here, we performed an integrative multi-omics analysis of patients diagnosed with breast cancer. Using transcriptomic analysis, we identified three subtypes (cluster A, cluster B and cluster C) of breast cancer with distinct prognosis, clinical features, and genomic alterations: Cluster A was associated with higher genomic instability, immune suppression and worst prognosis outcome; cluster B was associated with high activation of immune-pathway, increased mutations and middle prognosis outcome; cluster C was linked to Luminal A subtype patients, moderate immune cell infiltration and best prognosis outcome. Combination of the three newly identified clusters with PAM50 subtypes, we proposed potential new precision strategies for 15 subtypes using L1000 database. Then, we developed a robust gene pair (RGP) score for prognosis outcome prediction of patients with breast cancer. The RGP score is based on a novel gene-pairing approach to eliminate batch effects caused by differences in heterogeneous patient cohorts and transcriptomic data distributions, and it was validated in ten cohorts of patients with breast cancer. Finally, we developed a user-friendly web-tool (https://sujiezhulab.shinyapps.io/BRCA/) to predict subtype, treatment strategies and prognosis states for patients with breast cancer.
RÉSUMÉ
Ensemble learning is a kind of machine learning method which can integrate multiple basic learners together and achieve higher accuracy. Recently, single machine learning methods have been established to predict survival for patients with cancer. However, it still lacked a robust ensemble learning model with high accuracy to pick out patients with high risks. To achieve this, we proposed a novel genetic algorithm-aided three-stage ensemble learning method (3S score) for survival prediction. During the process of constructing the 3S score, double training sets were used to avoid over-fitting; the gene-pairing method was applied to reduce batch effect; a genetic algorithm was employed to select the best basic learner combination. When used to predict the survival state of glioma patients, this model achieved the highest C-index (0.697) as well as area under the receiver operating characteristic curve (ROC-AUCs) (first year = 0.705, third year = 0.825 and fifth year = 0.839) in the combined test set (n = 1191), compared with 12 other baseline models. Furthermore, the 3S score can distinguish survival significantly in eight cohorts among the total of nine independent test cohorts (P < 0.05), achieving significant improvement of ROC-AUCs. Notably, ablation experiments demonstrated that the gene-pairing method, double training sets and genetic algorithm make sure the robustness and effectiveness of the 3S score. The performance exploration on pan-cancer showed that the 3S score has excellent ability on survival prediction in five kinds of cancers, which was verified by Cox regression, survival curves and ROC curves together. To enable its clinical adoption, we implemented the 3S score and other two clinical factors as an easy-to-use web tool for risk scoring and therapy stratification in glioma patients.