FGFR Inhibitor Sensitivity in Cholangiocarcinoma with Interpretable Machine Learning

A composite FGFR inhibitor sensitivity score was derived from large-scale pharmacogenomic screening and baseline transcriptomes. The dataset includes a compact, interpretable transcriptional signature validated across multiple cholangiocarcinoma cohorts and an orthogonal resource (GDSC). Yading Xie authored this work, last updated on 2026-04-30.

Use Cases

• Predicting FGFR inhibitor sensitivity in cholangiocarcinoma based on baseline transcriptomic profiles.
• Validating pharmacogenomic models across platforms using matched cell line data from PRISM and GDSC.
• Extracting interpretable transcriptional programs from linear model coefficients for biological insight.
• Assessing biomarker reliability using a multi-compound, pathway-level phenotype to reduce single-drug bias.

Strengths

• The composite FGFRi score showed strong split-half reliability with a median ρ of 0.63.
• Methodology employed strict group-aware cross-validation to prevent information leakage.
• A reduced 15-feature panel preserved concordance with the full transcriptional signature.

Limitations

• Row count is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• Data may reflect bias inherent to the specific pharmacogenomic screening platforms used.

Provenance

Source

figshare, authored by Yading Xie.

Collection Method

Integrated large-scale pharmacogenomic screening with baseline transcriptomes, using cross-platform validation.

Freshness

Last updated 2026-04-30 05:24:00; freshness should be verified.