What question did this study set out to answer?

The research aims to leverage machine learning to develop scores for better prognosis and personalized therapy in ccRCC.

February 28, 2026Open Access

Machine learning-derived AS and AIS scores leverage BCAA metabolism and IL4I1 activity for prognosis and tailored therapy in ccRCC

Key Points

The research aims to leverage machine learning to develop scores for better prognosis and personalized therapy in ccRCC.
Utilized single-cell data analysis to examine branched-chain amino acid metabolism in ccRCC patients.
Developed Amino acid Signature Score (AS score) using ten machine learning algorithms.
Constructed an Amino acid Individualized Signature Score (AIS score) using principal component analysis.
Investigated the interaction between IL4I1 and branched-chain amino acid metabolism.
Branched-chain amino acid metabolism significantly impacts ccRCC progression and treatment.
The AS score accurately distinguishes drug sensitivities among patient subgroups.
The AIS score enhances outcomes for second-line and immunotherapy when targeted therapies fail.
Elevated IL4I1 expression correlates with increased tumor growth and metastasis.

Abstract

Background and objective Renal cell carcinoma (RCC) is among the most prevalent malignant tumors globally, characterized by a poor prognosis. The 5-year survival rate for advanced clear cell renal cell carcinoma (ccRCC) is below 20%. Materials and methods This study utilized single-cell data analysis to examine the differences in branched-chain amino acid metabolism among ccRCC patients. Ten machine learning algorithms were employed to develop Amino acid Signature Score (AS score), integrating data from TCGA and GEO cohorts. We compared and validated the clinical characteristics, molecular features, and drug sensitivity of patients with varying AS scores. To address patient heterogeneity, principal component analysis was applied to construct an Amino acid Individualized Signature Score (AIS score) aimed at guiding personalized treatment and assessing its performance in immunotherapy and targeted therapy. Additionally, we explored the interaction between IL4I1 and branched-chain amino acid metabolism, along with the underlying causes of abnormal expression, using spatial transcriptomics and single-cell multi-omics approaches. Results Branched-chain amino acid metabolism plays a crucial role in the progression and treatment of ccRCC. The AS score effectively distinguishes clinical characteristics and drug sensitivity across different patient subgroups. The AIS score confers a strategic advantage for second-line and immunotherapy when targeted therapy is ineffective. The elevated expression of IL4I1 enhances the degradation of branched-chain amino acids, promoting tumor growth and metastasis. Further analysis indicated that VHL mutations may elevate IL4I1 expression in tumors by modulating key transcription factors Hif-1a and SFMBT1, thus aggravating tumor progression. Conclusion Branched-chain amino acid metabolism and IL4I1 are pivotal in the progression of ccRCC. AS classification and the AIS score present a robust framework for personalized treatment strategies, while IL4I1 shows potential as a novel therapeutic target to enhance treatment efficacy.

Mark Helpful

Bookmark

Relay

View Full Paper