Integrative analysis with machine learning identifies diagnostic and prognostic signatures in neuroblastoma based on...

Area of Science:

• Oncology
• Genomics
• Bioinformatics

Background:

• Aberrant enhancer methylation is implicated in cancer gene expression.
• The role of differentially expressed enhancers in neuroblastoma (NB) stages 4S and 4 is not well understood.

Purpose of the Study:

• To identify differentially expressed enhancers between stage 4S and 4 NB.
• To develop a diagnostic and prognostic signature for NB using machine learning.
• To investigate the clinical and biological implications of the developed signature.

Main Methods:

• Utilized transcriptome and methylation data for Enhancer Linking by Methylation/Expression Relationships (ELMER) analysis.
• Established the INSS stage related signature (ISRS) by combining 22 machine learning algorithms.
• Validated ISRS performance, stratified patients into high/low-risk groups, and performed multi-omic comparisons.

Main Results:

• The ISRS accurately diagnosed stage 4 NB and predicted prognosis, outperforming existing signatures.
• Low-risk patients exhibited superior survival, enhanced immune infiltration, fewer mutations, and better immunotherapy response.
• Differential expression of CAMTA2 and FOXD1 was validated, correlating with prognosis.

Conclusions:

• The ISRS, developed using multi-omics data and machine learning, enables accurate NB diagnosis and prognostic stratification.
• The ISRS provides insights into NB molecular mechanisms, including spontaneous regression.
• The ISRS holds potential for clinical application in NB management.