Background: Medical texts such as radiology reports or electronic health records are a powerful source of data for researchers. Anonymization methods must be developed to de-identify documents containing personal information from both patients and medical staff. Although currently there are several anonymization strategies for the English language, they are also language-dependent. Here, we introduce a named entity recognition strategy for Spanish medical texts, translatable to other languages. Results: We tested 4 neural networks on our radiology reports dataset, achieving a recall of 97.18% of the identifying entities. Alongside, we developed a randomization algorithm to substitute the detected entities with new ones from the same category, making it virtually impossible to differentiate real data from synthetic data. The three best architectures were tested with the MEDDOCAN challenge dataset of electronic health records as an external test, achieving a recall of 69.18%. Conclusions: The strategy proposed, combining named entity recognition tasks with randomization of entities, is suitable for Spanish radiology reports. It does not require a big training corpus, thus it could be easily extended to other languages and medical texts, such as electronic health records.

Pancreatic ductal adenocarcinoma (PDAC) prognoses and treatment responses remain devastatingly poor due partly to the highly heterogeneous, aggressive, and immunosuppressive nature of this tumor type. The intricate relationship between the stroma, inflammation, and immunity remains vaguely understood in the PDAC microenvironment. Here, we performed a meta-analysis of stroma-, and immune-related gene expression in the PDAC microenvironment to improve disease prognosis and therapeutic development. We selected 21 PDAC studies from the Gene Expression Omnibus and ArrayExpress databases, including 922 samples (320 controls and 602 cases). Differential gene enrichment analysis identified 1153 significant dysregulated genes in PDAC patients that contribute to a desmoplastic stroma and an immunosuppressive environment (the hallmarks of PDAC tumors). The results highlighted two gene signatures related to the immune and stromal environments that cluster PDAC patients into high- and low-risk groups, impacting patients’ stratification and therapeutic decision making. Moreover, HCP5, SLFN13, IRF9, IFIT2, and IFI35 immune genes are related to the prognosis of PDAC patients for the first time.

Pancreatic ductal adenocarcinoma (PDAC) prognosis and treatment response remains devastatingly poor due partly to the highly heterogeneous, aggressive, and immunosuppressive nature of this tumor type. The intricate relationship between stroma, inflammation, and immunity remains vaguely understood in the PDAC microenvironment. Here, we performed a meta-analysis of stroma-, and immune-related gene expression in the PDAC microenvironment to improve disease prognosis and therapeutic development. We selected twenty-one PDAC studies from the Gene Expression Omnibus and ArrayExpress databases, including 922 samples (320 controls and 602 cases). Differential gene enrichment analysis identified 1153 significant dysregulated genes in PDAC patients that contribute to a desmoplastic stroma and an immunosuppressive environment (the hallmarks of PDAC tumors). The results highlighted two gene signatures related to the immune and stromal environments that cluster PDAC patients in high- and low-risk groups, impacting patient stratification and therapeutic decision-making. Moreover, HCP5, SLFN13, IRF9, IFIT2 , and IFI35 immune genes were related to prognosis value in PDAC patients, for the first time. Simple Summary: Pancreatic ductal adenocarcinoma (PDAC) is a highly lethal disease with few curative options. Desmoplastic stroma and immune system evasion in PDAC represent challenges to the success of therapeutic strategies that function well in other tumor types. Characterizing the PDAC microenvironment (including the immune environment) remains critical to developing safe and efficient therapies. Here, we present a comprehensive meta-analysis identifying 1153 significantly dysregulated genes, which mainly impact extracellular matrix remodeling and the immune system. We identify two signatures of twenty-eight immune-related genes and eleven stroma-related genes influencing PDAC patient survival. Additionally, five immune genes are associated with PDAC prognosis for the first time.

Medical texts such as radiology reports or electronic health records are a powerful source of data for researchers. Anonymization methods must be developed to de-identify documents containing personal information from both patients and medical staff. Although currently there are several anonymization strategies for the English language, they are also language-dependent. Here, we introduce a named entity recognition strategy for Spanish medical texts, translatable to other languages. We tested 4 neural networks on our radiology reports dataset, achieving a recall of 97.18% of the identifying entities. Along-side, we developed a randomization algorithm to substitute the detected entities with new ones from the same category, making it virtually impossible to differentiate real data from synthetic data. The three best architectures were tested with the MEDDOCAN challenge dataset of electronic health records as an external test, achieving a recall of 69.18%. The strategy proposed, combining named entity recognition tasks with randomization of entities, is suitable for Spanish radiology reports. It does not require a big training corpus, thus it can be easily extended to other languages and medical texts, such as electronic health records.

While studies have established the existence of differences in the epidemiological and clinical patterns of lung adenocarcinoma between male and female patients, we know relatively little regarding the molecular mechanisms underlying such sex-based differences. In this study, we explore said differences through a meta-analysis of transcriptomic data. We performed a meta-analysis of the functional profiling of nine public datasets that included 1,366 samples from Gene Expression Omnibus and The Cancer Genome Atlas databases. Meta-analysis results show an enrichment of Gene Ontology terms and Kyoto Encyclopedia of Genes and Genomes pathways related to the immune response, nucleic acid metabolism, and purinergic signaling. We discovered the overrepresentation of terms associated with the immune response, particularly with acute inflammatory response, and purinergic signaling in female lung adenocarcinoma patients, which could influence reported clinical differences. Further evaluations of the identified differential biological processes and pathways could lead to the discovery of new biomarkers and therapeutic targets. Our findings also emphasize the relevance of sex-specific analyses in biomedicine, which represents a crucial aspect influencing biological variability in disease.

Summary Sex and gender differences in different health scenarios has been thoroughly acknowledged in the literature, and yet, very scarcely analyzed. To fill the gap, here we present MetaFun, which allows to meta-analyze multiple omics datasets with a sex-based perspective, and to combine different datasets to gain major statistical power and to assist the researcher in understanding these sex differences in the diseases under study. Metafun is freely available at bioinfo.cipf.es/metafun Availability and implementation: MetaFun is available under http://bioinfo.cipf.es/metafun. The backend has been implemented in R and Java and the frontend has been developed using Angular. Supplementary information R code available at https://gitlab.com/ubb-cipf/metafunr