Decoding the JAK-STAT Axis in Colorectal Cancer with AI-HOPE-JAK-STAT: A Conversational Artificial Intelligence...

Area of Science:

• Oncology
• Bioinformatics
• Artificial Intelligence

Background:

• The Janus kinase-signal transducer and activator of transcription (JAK-STAT) pathway is crucial in immune regulation, inflammation, and cancer.
• Its role in colorectal cancer (CRC) pathogenesis, especially in early-onset CRC (EOCRC), requires further characterization.
• Molecular heterogeneity and clinical significance of JAK-STAT alterations in CRC are not fully understood across diverse contexts.

Purpose of the Study:

• To introduce AI-HOPE-JAK-STAT, a novel conversational AI platform for real-time, natural language-driven exploration of JAK/STAT pathway alterations in CRC.
• To integrate clinical, genomic, and treatment data for dynamic, hypothesis-generating analyses in precision oncology.
• To enable users to explore JAK/STAT pathway alterations without requiring coding expertise.

Main Methods:

• Development of AI-HOPE-JAK-STAT, combining large language models (LLMs) and a natural language-to-code engine.
• Integration of harmonized public CRC datasets from cBioPortal for cohort selection, survival analysis, and mutation profiling.
• Validation of the platform by replicating known JAK1, JAK3, and STAT3 mutation associations and conducting exploratory analyses on age, treatment, stage, and site.

Main Results:

• The platform confirmed improved survival for EOCRC patients with JAK/STAT pathway alterations.
• JAK/STAT-altered tumors in FOLFOX-treated CRC cohorts showed significantly enhanced overall survival (p < 0.0001).
• Younger patients (age < 50) with JAK/STAT mutations demonstrated survival advantages (p = 0.0379), and STAT5B mutations correlated with favorable trends (p = 0.0000).
• JAK1 mutations in microsatellite-stable tumors did not impact survival, highlighting the importance of molecular context.
• JAK3-mutated tumors in Stages I-III had superior survival compared to Stage IV (p = 0.00001), underscoring stage as a key determinant.

Conclusions:

• AI-HOPE-JAK-STAT sets a new standard for pathway-level analysis in CRC, empowering hypothesis generation and testing.
• The system improves access to precision oncology analyses, facilitating scalable, real-time discovery of survival trends and treatment-response patterns.
• It supports the identification of mutational associations and survival trends across stratified patient cohorts without coding expertise.