Fig. 2: External validation of molecular subtypes PACpAInt-B/C and association with survival. | Nature Communications

Fig. 2: External validation of molecular subtypes PACpAInt-B/C and association with survival.

From: Pacpaint: a histology-based deep learning model uncovers the extensive intratumor molecular heterogeneity of pancreatic adenocarcinoma

Fig. 2

a Simplified workflow of the study: a first model is applied to find the tumor (PACpAInt-Neo) followed by a second model predicting the global tumor cell molecular type (classical vs basal-like) at the slide level (PACpAInt-B/C), b Representative tiles identified as classical or basal-like by PACpAInt-B/C in the validation BJN_U cohort (112 μm square), c–e Performance of PACpAInt-B/C to identify molecular subtypes using the whole cohort or only cases with an unambiguous RNA subtype (clear subtype) of the validation cohorts (BJN_U unmatched (surgical specimens), i.e., slide analyzed and tissue used for RNAseq are not spatially matched; BJN_M matched (surgical specimens), i.e., slide analyzed and tissue used for RNAseq are spatially matched; Liver_FNB (EUS fine-needle biopsies), f, g Multivariate analyses of clinical/pathological factors and PACpAInt-B/C demonstrating an independent prognostic value of the later on disease-free survival (n = 243) and overall survival (n = 248). The circle represents the variable hazard ratio, while whiskers represent the 95% confidence interval of that hazard ratio. P values were computed using a two-sided Wald test. No adjustments for multiple comparisons were made. ***P < 0.001; **P < 0.01; *P < 0.05; +P < 0.1; −P > 0.1. Source data are provided as a Source Data file.

Back to article page