Q-omics provides the consensus-scored HLA-G profile across patient tissues and cancer cell-line models. HLA-G expression is associated with patient survival in 25 of 34 cancer types, with the highest sampling consensus in UVM. Among the 18 cancer types available for tumor–normal comparison, HLA-G is differentially expressed in 11, with the highest sampling consensus in KIRC. Additionally, HLA-G RNA expression shows 12,443 significant gene co-expression associations, with the highest sampling consensus in UVM. Together, these results highlight UVM, and KIRC as cancer lineages where HLA-G shows reproducible signals across survival, tumor–normal expression, and patient cross-omics analyses.
Every result is evaluated using two consensus scores. Sampling consensus measures how consistently a finding is reproduced within a cancer lineage across different conditions. Lineage consensus measures how broadly the result is shared across cancer types, distinguishing pan-cancer signals from lineage-specific patterns.
Premium analyses for HLA-G — synthetic lethality, tumor antigen, and pembrolizumab response.
This table summarizes HLA-G survival associations across molecular data types. HLA-G RNA expression shows survival associations in the most cancer types (25), followed by mutation status (6) and mass-spec protein abundance (1). The rightmost column indicates the cancer type with the highest sampling consensus for each molecular layer.
This table ranks reproducible HLA-G RNA expression–survival associations across cancer types. High HLA-G expression shows unfavorable associations in UVM, LGG and STAD, but favorable associations in SKCM, BLCA and SCLC. The UVM Kaplan–Meier curve shows clear separation, with the high-expression group declining faster, consistent with the unfavorable association (log-rank p < 0.001). Together, the overview and detailed table identify UVM as the clearest survival context for HLA-G RNA expression.
This table summarizes HLA-G tumor–normal expression differences by data type. RNA shows broader differences across cancer types, with a lineage consensus of 11, while mass-spec protein shows differences in 3. The strongest signals are observed in KIRC for RNA and CCRCC for protein.
This table ranks reproducible tumor–normal expression differences for HLA-G. A negative fold-change indicates higher expression in normal tissue than in tumor tissue. HLA-G shows lower tumor expression in LUSC and higher tumor expression in KIRC, HNSC, THCA, KIRP and BRCA. The KIRC box plot shows higher HLA-G RNA expression in tumor versus normal tissue (log2 FC = +3.046, t-test p < 0.001).
This table shows molecular features associated with HLA-G in patient tissues and cancer cell lines. In patient samples, HLA-G shows the broadest associations at the RNA and protein expression levels, with UVM recurring as the lineage with the largest associated feature set. In cancer cell lines, HLA-G RNA and mutation anchors are most strongly linked to RNA-expression features, especially in BLOOD_Myeloma, while CRISPR and shRNA rows add functional-dependency signals in BLOOD_Leukemia and CNS.
