Development and validation of a kidney renal clear cell carcinoma prognostic model relying on pyroptosis-related LncRNAs-A multidimensional comprehensive bioinformatics exploration

Abstract

Background: Renal cell carcinoma (RCC) is a malignant tumour that may develop in the kidney. RCC is one of the most common kinds of tumours of this sort, and its most common pathological subtype is kidney renal clear cell carcinoma (KIRC). However, the aetiology and pathogenesis of RCC still need to be clarified. Exploring the internal mechanism of RCC contributes to diagnosing and treating this disease. Pyroptosis is a critical process related to cell death. Recent research has shown that pyroptosis is a critical factor in the initiation and progression of tumour formation. Thus far, researchers have progressively uncovered evidence of the regulatory influence that long noncoding RNAs (lncRNAs) have on pyroptosis.

Methods: In this work, a comprehensive bioinformatics approach was used to produce a predictive model according to pyroptosis-interrelated lncRNAs for the purpose of predicting the overall survival and molecular immune specialties of patients diagnosed with KIRC. This model was verified from multiple perspectives.

Results: First, we discovered pyroptosis-associated lncRNAs in KIRC patients using the TCGA database and a Sankey diagram. Then, we developed and validated a KIRC patient risk model based on pyroptosis-related lncRNAs. We demonstrated the grouping power of PLnRM through PCA and used PLnRM to assess the tumour immune microenvironment and response to immunotherapy. Immunological and molecular traits of diverse PLnRM subgroups were evaluated, as were clinical KIRC patient characteristics and predictive risk models. On this basis, a predictive nomogram was developed and analyzed, and novel PLnRM candidate compounds were identified. Finally, we investigated possible medications used by KIRC patients.

Conclusions: The results demonstrate that the model generated has significant value for KIRC in clinical practice.

Keywords: Comprehensive bioinformatics; Immune response; Kidney renal clear cell carcinoma; Long noncoding RNA; Pyroptosis.

Fig. 1

Comprehensive research workflow

Fig. 2

Discovery of lncRNAs associated with pyroptosis in KIRC patients. A Fifty-two pyroptosis genes and lncRNAs are represented by a Sankey diagram. B Heatmap depicting the relationship between 52 pyroptosis genes and 6 prognostic pyroptosis-associated lncRNAs

Fig. 3

The risk model for KIRC patients based on pyroptosis-related lncRNAs. A The identified lncRNAs have a substantial correlation with clinical prognosis, as indicated by univariate Cox regression analysis. B The LASSO coefficient profile of ten OS-related lncRNAs and imaginary perpendicular lines were drawn at the value determined by tenfold cross-validation. C Tuning parameters (logλ) of OS-related proteins were selected for error curve cross-validation. At the ideal value, perpendicular fictitious lines were drawn in accordance with the minimum criteria and 1-se criterion. D–F OS Kaplan‒Meier survival curves for high- and low-risk patient populations (the entire TCGA, training, and test sets)

Fig. 4

Assessment of the predictive value of PLnRM risk patterns in the TCGA dataset. A1 Distribution of PLnRM-based risk scores. A2 Both groups exhibit distinct survival status and duration patterns. A3 The heatmap resulting from clustering analysis displays expression levels of PLnRM in each patient. B–C Relevant findings from both the training and test sets

Fig. 5

Independent cohort validation. A OS Kaplan‒Meier curves for high-risk and low-risk patient groups. B Model-based risk score distribution for m6A-related lncRNAs in the validation set. C ROC curves for clinical characteristics (3 and 5 years)

Fig. 6

We conducted principal component analysis on four different datasets: A entire gene expression profiles, B 52 pyroptosis genes, C 576 lncRNAs associated with pyroptosis, and D a risk model using six pyroptosis-related long noncoding RNAs from the entire TCGA dataset

Fig. 7

Estimation of the tumour immune microenvironment and immunotherapy response using PLnRM for the full TCGA dataset. A GO enrichment analysis. B The specified requirements for each patient's immunity index. C, D The waterfall plot shows gene mutation data for high-risk and low-risk groups, specifically highlighting genes with high mutation rates. E TMB was different between the patients in the two groups. F–G Patient mutation status (high or low) and their PLnRM were taken into account for the Kaplan‒Meier analysis of their OS. H Differences in TIDE prediction for patients between the two groups

Fig. 8

Molecular and immune features of several PLnRM subgroups. A A bar graph depicting the relative proportion of 21 immune cells infiltrating tumours in the high-risk and low-risk groups. B The violin plot illustrates the disparity in the proportions of each type of immune cell in the two risk groups. C Qualities associated with the immune system. D and E Gene set enrichment analysis using the Kyoto Encyclopedia of Genes and Genomes (KEGG) for the low-risk group and high-risk group. F–J The two groups differ in expression and associations of common immunological checkpoints

Fig. 9

Evaluation of the clinical characteristics of KIRC as well as the prognostic risk model using the whole TCGA dataset. A–B Assessments of the clinical features and risk scores linked with OS using both univariate and multivariate methods. C Concordance indicators for the risk score and clinical features. D–E ROC curves of various clinical parameters and risk scores

Fig. 10

A predictive nomogram was constructed and evaluated. A The nomogram predicts the probability of 1-, 2-, and 3-year overall survival. B The nomogram calibration plot was created using the OS probabilities for 1, 2, and 3 years