Predicting neuroblastoma using developmental signals and a logic-based model.

Area of Science:

• Oncology
• Systems Biology
• Computational Biology

Background:

• Current neuroblastoma progression prediction relies on gene expression correlations, lacking mechanistic insight.
• Existing methods show limited potential for accurate disease outcome prediction.
• A need exists for robust molecular network models in neuroblastoma research.

Purpose of the Study:

• To construct and simulate a 6-gene logic-based molecular network model for predicting neuroblastoma progression.
• To investigate the impact of tyrosine receptor kinase mis-expression on neuroblastoma outcomes.
• To compare the predictive accuracy of the developed model against existing gene lists.

Main Methods:

• A 6-gene input logic model simulating cell differentiation, proliferation, apoptosis, and angiogenesis was constructed.
• Simulations included mis-expression of tyrosine receptor kinases (trkA and trkB).
• Model assumptions were validated using RNAseq of the SHSY5Y cell line and antibody staining, and applied to 77 patient samples.

Main Results:

• The model successfully predicted favorable/unfavorable outcomes based on simulated cell states.
• Mis-expression of trkA and trkB led to distinct steady-state outcome distributions.
• The developed model demonstrated higher accuracy in predicting early-stage disease outcomes compared to current neuroblastoma gene lists.

Conclusions:

• Logic-based models utilizing developmental genes offer significant predictive power for neuroblastoma progression.
• The study enhances understanding of molecular network interactions in neuroblastoma.
• This approach provides a more accurate tool for predicting pediatric neuroblastoma patient outcomes.