Gastric poorly cohesive carcinoma (PCC) manifests with a diffuse pattern and diverse tumor cell morphologies, often indicating a more unfavorable prognosis. Recent consensus has reclassified PCC based on the proportion of signet-ring cells (SRCs) in tumors for research purposes. The two most distinct subtypes, poorly cohesive carcinoma not otherwise specified (PCC-NOS) and signet-ring cell carcinoma (SRCC), are characterized by less than 10% and more than 90% SRCs, respectively. However, research comparing the clinicopathological and transcriptomic differences between these subtypes remains limited. In this study, we conducted a comparative analysis of clinicopathological features in 55 advanced-stage PCCs, consisting of 43 PCC-NOS and 12 SRCC cases. Subsequently, 12 PCC-NOS and 5 SRCC cases were randomly selected for initial cancer-related gene expression profiling and pathway enrichment analysis using the GeoMx digital spatial profiler, followed by validation in a separate validation group comprising 16 PCC-NOS and 6 SRCC cases. These transcriptomic findings were then correlated with tumor morphology and clinicopathological data. PCC-NOS cases exhibited larger tumor size, a higher prevalence of pathological N3 disease, and a worse 1-year progression-free survival rate compared to SRCC cases. Clustering of PCC-NOS and SRCC was successfully achieved using the GeoMx Cancer Transcriptome Atlas. Among all studied genes, only MMP7 showed differential expression, with its overexpression significantly associated with the PCC-NOS subtype, increased perineural invasion, and earlier disease progression. Pathway analysis revealed significantly enriched pathways in PCC-NOS related to vesicle-mediated transport, adaptive immune systems, oncogenic signaling, and extracellular matrix organization, while SRCC displayed significant enrichment in pathways associated with respiratory electron transport and the cell cycle. In conclusion, this study compares and correlates clinicopathological features and transcriptomic data between PCC-NOS and SRCC at advanced stages, employing the latest consensus classification and a novel platform for analysis.