Page 58 - GPD-2-1

Gene & Protein in Disease Pyroptosis-related LncRNAs in pediatric AML

ESTIMATE and cell type identification by estimating clinical variables. Independent factors in the prognosis
relative subsets of RNA transcription (CIBERSORT) of pediatric AML patients were identified through uni-
algorithms , respectively. Furthermore, the correlations Cox and multi-Cox analyses. Stratification analyses were
[16]
among different subtypes in the expression of five immune performed to determine the stability of each clinical factor.
checkpoints, including programmed cell death protein The semi-inhibitory concentration (IC50) values of
1 (PD-1), programmed cell death-ligand 1 (PD-L1) , chemotherapeutic drugs that are generally used to treat
[17]
cytotoxic T-lymphocyte-associated protein 4 (CTLA-4) , AML were estimated by the “prophetic” package in R .
[18]
[21]
lymphocyte-activation gene 3 (LAG3) , and hepatitis A Besides, the “PreMSIm” package was used to predict the
[19]
virus cellular receptor 2 (HAVCR2/TIM-3) pathways, microsatellite instability (MSI) state in both high- and low-
[20]
which are implicated in tumor immune evasion and risk groups based on the 15 genes expression.
derived from previous studies, were analyzed.
Furthermore, Gene Ontology (GO) and KEGG
2.4. Differentially expressed lncRNAs recognition enrichment analyses were performed to determine the
and prognosis and pyroptosis-related lncRNAs function of the differentially expressed genes (DEGs)
signature construction between the two groups. The DEGs were screened with
Differently expressed lncRNAs (DE-lncRNAs) within the |log2FC| ≥ 1 and false discovery rate (FDR) < 0.05.
three clusters were identified by the “limma” package in R 2.6. Establishments of a nomogram and a decision
according to the following criteria: log2FC ≥ 1 and adjusted curve
P < 0.001. Uni-Cox was used to filter for DE-lncRNAs
according to prognosis. These lncRNAs were then used Combining the signature with clinical factors, a nomogram
to form PPR-lncRNAs signaling to predict the prognosis was constructed, integrating the prognostic signature using
of pediatric AML patients. First, 1300 pediatric AML the “rms” package in R, to predict the 1-, 3-, and 5-year
samples were randomly sorted into training and testing survival probability of pediatric AML patients. In addition,
sets at a ratio of 7:3. LASSO‐Cox ten-fold cross-validation a decision curve analysis (DCA) was used to calculate the
and multivariate Cox regression analysis (multi-Cox) net benefit of each factor on the survival of pediatric AML
were used to establish the PPR-lncRNAs signature in the patients at 1, 3, and 5 years.
training set. The formula for calculating the risk score is 2.7. Statistical analysis
shown below:
In our study, statistical analyses were performed using R
n software (version 4.1.2). If not specifically stated, all results
riskscore = ∑ coef * x i
i
i= 1 were regarded as statistically significant when P < 0.05.
Where coef represents the coefficients and x represents 3. Results
i
i
the count of PPR-lncRNAs expressions. Based on the Figure 1. Flow chart of 1300 samples with complete clinical
calculation of the risk score, the pediatric AML samples data from the TARGET database. Following Pearson
in the training and testing sets were divided into high- and correlation analysis and uni-Cox, 841 prognosis, and
low-risk groups based on the median risk score of the pyroptosis-related lncRNAs were obtained. Three clusters
training set. were classified by consensus clustering according to
pyroptosis-related lncRNAs. Based on these three clusters,
2.5. Validation of the signature
prognostic signature construction and immune difference
K-M curves, ESTIMATE and CIBERSORT scores, and exploration were performed.
immune checkpoint expression were used to assess the Table 1 shows the characteristics of 1300 pediatric
differences between the two groups. Moreover, time- AML patients from the TARGET database. All 1300
dependent receiver operating characteristic (ROC) curves AML patients with their OS information were used for
were used to assess the predictive ability of the prognostic prognostic model construction. From the expression
signature for OS.
matrix of 11,535 lncRNAs and 52 pyroptosis-related
Subgroup analyses of the selected clinical characteristics genes (PRGs), we identified 1792 lncRNAs as significant
(age, gender, race, bone marrow leukemic blast percentage pyroptosis-associated genes by Pearson (Table S1).
[BM], peripheral blasts [PB], white blood cell at diagnosis Three clusters were classified by unsupervised consensus
[WB], and French-American-British [FAB] category) clustering analysis and uni-Cox based on PR-lncRNAs
were performed. Chi-squared (χ ) test was performed to (Figure 2A-C). The age, gender, and race components
2
evaluate the distribution among subtypes, risk scores, and did not show any statistical difference (P > 0.05). Three-

Volume 2 Issue 1 (2023) 3 https://doi.org/10.36922/gpd.v2i1.230