ILnc: Prioritizing Long Non-coding RNAs for Pan-cancer Analysis of Immune Cell Infiltration

ABSTRACT

Purpose: The distribution and extent of immune cell infiltration into solid tumors play pivotal roles in cancer immunology and therapy. Here we introduced an immune long non-coding RNA (lncRNA) signature-based method (ILnc), for estimating the abundance of 14 immune cell types from lncRNA transcriptome data. Performance evaluation through pure immune cell data showed that our lncRNA signature sets were more accurate than protein- coding gene signatures. We found that lncRNA signatures were significantly enriched to immune functions and pathways, such as immune response and T cell activation. In addition, the expression of these lncRNAs was significantly correlated with expression of marker genes in corresponding immune cells. Methods: Application of ILnc in 33 cancer types provided a global view of immune infiltration across cancers and we found that the abundance of most immune cells is significantly associated with patient clinical signatures. Finally, six immune subtypes spanning cancer tissue types were identified, characterized by differences in immune cell infiltration, homologous recombination deficiency (HRD), expression of immune checkpoint genes, and prognosis. Results: Altogether, these results demonstrate that ILnc is a powerful and exhibits broad utility for cancer researchers to estimate tumor immune infiltration, which will be a valuable tool for precise classification and clinical prediction. has obtained great successes in the clinical application,