Data Sheet 2_A postoperative recurrence prediction model for intrahepatic cholangiocarcinoma based on multi-omics analysis of adjacent-to-tumor tissues.docx

Rongyu Ping's research provides a multivariate Cox regression model for predicting postoperative recurrence of intrahepatic cholangiocarcinoma (iCCA). The model is based on proteomic data from 116 iCCA adjacent-to-tumor tissue (ATT) samples and validated on an independent cohort of 88 samples. It identifies four key molecular features (ENO3, HSPA13, POSTN, PTBP3) and stratifies patients into high-risk and low-risk recurrence subtypes.

Use Cases

• Train a survival model using the four-protein signature (ENO3, HSPA13, POSTN, PTBP3) to predict recurrence-free survival for iCCA patients.
• Validate the POSTN expression level as an independent prognostic factor for recurrence using immunohistochemistry data from the 88-sample validation cohort.
• Analyze the association between the S1/S2 molecular subtypes derived from ATT proteomics and clinical outcomes like recurrence rates and immune scores.
• Investigate treatment response correlations by assessing whether the high-risk recurrence group shows differential sensitivity to immunotherapy, chemotherapy, or targeted therapies.

Strengths

• Model built on proteomic data from 116 iCCA patient samples, providing a substantive biological basis.
• Includes independent validation on a separate cohort of 88 iCCA tissue samples via tissue microarray and immunohistochemistry.
• Identifies a specific four-molecule signature (ENO3, HSPA13, POSTN, PTBP3) with POSTN confirmed as an independent prognostic factor.

Limitations

• Sample size is moderate; the primary cohort has 116 patients and the validation cohort has 88, which may limit statistical power for rare subgroups.
• The underlying proteomic and validation data are not directly provided in a structured, machine-readable format (e.g., CSV); the primary file is a 17.5 MB DOCX document.
• Clinical and demographic details of the patient cohorts are not specified, potentially limiting the assessment of population representativeness.

Provenance

Source

figshare, authored by Rongyu Ping.

Collection Method

Proteomic analysis of adjacent-to-tumor tissues from iCCA patients, consensus clustering, multivariate Cox regression, and validation via tissue microarray immunohistochemistry.

Freshness

Last updated March 25, 2026.