Pan-cancer analysis of clinical significance and associated molecular features of glycolysis

Figures & data

Table 1. Univariate cox analysis of glycolysis score

Study	N	Hazard ratio	95% confidence interval	P-value
BLCA	399	1.793	0.954–3.367	0.070
BRCA	1052	1.894	0.974–3.684	0.060
CESC	273	2.591	0.931–7.214	0.068
COAD	426	0.672	0.277–1.631	0.380
ESCA	178	0.985	0.404–2.399	0.973
GBM	146	1.034	0.400–2.676	0.945
HNSC	512	2.819	1.553–5.117	0.001
KIRC	518	0.972	0.460–2.053	0.940
KIRP	278	3.941	0.840–18.483	0.082
LGG	481	3.276	1.161–9.243	0.025
LIHC	343	8.862	3.471–22.628	5.00E-06
LUAD	492	3.460	1.940–6.173	2.60E-05
LUSC	474	0.714	0.401–1.271	0.252
OV	295	0.724	0.316–1.658	0.444
PAAD	172	3.034	1.113–8.271	0.030
PCPG	172	10.703	0.112–1020.485	0.308
PRAD	494	20.645	0.705–604.846	0.079
READ	153	0.243	0.042–1.403	0.114
SARC	255	3.723	1.571–8.825	0.003
STAD	375	0.609	0.341–1.087	0.093
TGCT	130	0.787	0.015–40.206	0.905
THCA	502	4.915	0.402–60.072	0.213
THYM	118	0.112	0.008–1.580	0.105
UCEC	513	1.389	0.622–3.100	0.423

Table 2. Subgroup analysis for the prognostic value of glycolysis score

Parameters	Types of cancer	No. of patients	HR (95%CI)	P-value	Model
Age
Age <60	23	4034	1.82 (1.08–3.08)	0.025	Random effects
Age ≥60	22	4575	1.50 (1.01–2.22)	0.044	Random effects
Gender
Female	21	4477	1.62 (1.25–2.11)	<0.001	Fixed effects
Male	20	4156	1.81 (1.11–2.93)	0.016	Random effects
Stage
I/II	15	3631	1.82 (1.07–3.08)	0.027	Random effects
III/IV	14	2078	1.56 (1.17–2.10)	0.003	Fixed effects

Figure 1. The prognostic value of glycolysis in pan-cancer patients

(A) Forest plot of the correlation between glycolysis levels and OS of pan-cancer patients; (B) Heatmap shows P-values (including age, gender, pathologic stage, and histological grade) with significance (P < 0.05, red); (C–F) Highlight of tumor molecular subtypes that exhibit strong correlations with glycolysis.

Figure 2. Validation of the prognostic value of hepatocellular carcinoma (HCC) in three independent cohorts

(A) GSE14520; (B) GSE54236; (C) LIRI-JP.

Figure 3. Exploration of metabolism-driven cancer types

(A) Patients were divided into high- and low- glycolysis groups based on the gene expression of 72 signature genes generated by K-means clustering; (B) Principal component analysis plot indicates that the two subgroups have distinct glycolysis gene expression profiles; (C) Kaplan–Meier survival analysis curves for the two patient clusters; (D) Heatmap of the 8633 patients grouped by cluster, with annotations associated with each cluster; (E) The ratio of patients with high and low glycolysis in different cancer types.

Figure 4. Associations between glycolysis and 10 oncogenic signaling pathway alterations

(A) Heatmap of glycolysis scores and pathway alterations; (B) Heatmap showing P-values (Mann–Whitney U test). Red indicates that the glycolysis score is upregulated, while blue indicates that it is downregulated in the pathway alteration group.

Figure 5. Gene functional enrichment analysis of glycolysis-associated genes

(A) Gene ontology of glycolysis-associated genes; (B) Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis of glycolysis-associated genes; (C) Protein–protein interaction (PPI) network of glycolysis-associated genes.

Figure 6. Correlation network of glycolysis-related proteins

(A) Red lines show proteins positively related to glycolysis, while blue lines show proteins negatively related to glycolysis; (B) Bar plot indicates that cyclin B1 positively correlates with glycolysis in 16 types of cancer.

Figure 7. Associations between glycolysis and immune cell infiltrations

Figure 8. Correlation of glycolysis with drug resistance: connectivity map analysis

(A) Heatmap showing a negative enrichment score of each compound from Cmap for each cancer type; (B) Heatmap showing each Cmap compound (perturbagen) that shares mechanisms of action (rows) and sorted by descending number of compounds with shared mechanisms of action.

Data availability statement

The datasets analyzed in this study were obtained from the TCGA database (http://www.cancer.gov/tcga).