Fig. 5: GLASS-AI identifies altered Mek/Erk signaling in high-grade tumor regions. | npj Precision Oncology

Fig. 5: GLASS-AI identifies altered Mek/Erk signaling in high-grade tumor regions.

From: Grading of lung adenocarcinomas with simultaneous segmentation by artificial intelligence (GLASS-AI)

Fig. 5

a The proportion of LUAD tumors of each grade that were positively stained for p-Mek or p-Erk was determined for each of the K, TK, and KP-R172H LUAD mouse models. Data represent mean ± SD. b Overlays of IHC stain analysis and tumor grading from GLASS-AI were generated after registration to adjacent H&E slides. Original magnification ×20, scale bars represent 100 µm. c, d Positively stained tumors were tested for uneven distribution of p-Mek (c) or p-Erk staining (d) using a likelihood-ratio G-test. Significance was determined by p < 0.01 versus a Chi-squared distribution. e Localization of p-Mek and p-Erk enrichment was evaluated using individual likelihood ratios for each grade region within individual tumors that displayed uneven staining distribution of both p-Mek and p-Erk in panels (c, d). Individual tumors are matched vertically across the three subgraphs. Colors represent the likelihood ratio of each IHC stain within the region of the indicated grade superimposed together (top and bottom) or the percentage of the tumors’ area comprised of each tumor grade (middle). Black dots indicate the overall tumor grade based on the highest tumor grade present that comprises ≥10% of the tumor’s total area (middle).

Back to article page