Integrated bioinformatics analysis of differentially expressed genes and immune cell infiltration characteristics in Esophageal Squamous cell carcinoma

Abstract

Esophageal squamous cell carcinoma (ESCC) is a life-threatening thoracic tumor with a poor prognosis. The role of molecular alterations and the immune microenvironment in ESCC development has not been fully elucidated. The present study aimed to elucidate key candidate genes and immune cell infiltration characteristics in ESCC by integrated bioinformatics analysis. Nine gene expression datasets from the Gene Expression Omnibus (GEO) database were analysed to identify robust differentially expressed genes (DEGs) using the robust rank aggregation (RRA) algorithm. Functional enrichment analyses showed that the 152 robust DEGs are involved in multiple processes in the tumor microenvironment (TME). Immune cell infiltration analysis based on the 9 normalized GEO microarray datasets was conducted with the CIBERSORT algorithm. The changes in macrophages between ESCC and normal tissues were particularly obvious. In ESCC tissues, M0 and M1 macrophages were increased dramatically, while M2 macrophages were decreased. A robust DEG-based protein–protein interaction (PPI) network was used for hub gene selection with the CytoHubba plugin in Cytoscape. Nine hub genes (CDA, CXCL1, IGFBP3, MMP3, MMP11, PLAU, SERPINE1, SPP1 and VCAN) had high diagnostic efficiency for ESCC according to receiver operating characteristic (ROC) curve analysis. The expression of all hub genes except MMP3 and PLAU was significantly related to macrophage infiltration. Univariate and multivariate regression analyses showed that a 7-gene signature constructed from the robust DEGs was useful for predicting ESCC prognosis. Our results might facilitate the exploration of potential targeted TME therapies and prognostic evaluation in ESCC.

Radiomics-based machine learning differentiates “ground-glass” opacities due to COVID-19 from acute non-COVID-19 lung disease

“…Our results demonstrated that a machine learning signature based on radiomics features extracted from GGOs on CT images is an accurate method to early differentiate COVID-19 pneumonia from other acute non-COVID-19 lung diseases. These results confirm the promising role of radiomics in the diagnosis of COVID-19 pneumonia and are in line with the most recent literature on this topic 21 – 24 . For example, Huang et al used radiomics to discriminate COVID-19 and influenza pneumonia by combining CT signs and quantitative features extracted from the initial unenhanced CT images 22 .…”

“…For this reason, although the context of the current pandemic and the previous CT findings can be indicative of COVID-19 pneumonia, the differential diagnosis of GGOs remains a challenge 10 , 17 . Recently, a few studies outlined a growing interest toward imaging-based tools aimed to assist physicians in patient management 18 – 24 . Radiomics is one of these tools and it allows the extraction of large amounts of quantitative data from medical images 25 .…”

“…Radiomics features can be further integrated into machine learning models with the aim to improve diagnosis and patient management. This approach was recently investigated to improve the detection and the differential diagnosis of COVID-19 pneumonia 20 , 21 , 23 , 24 , 26 – 29 . For example, Zhang et al proposed a CT-based deep learning integrated radiomics model for the differentiation of COVID-19 pneumonia from other community acquired pneumonias.…”

“…For example, Zhang et al proposed a CT-based deep learning integrated radiomics model for the differentiation of COVID-19 pneumonia from other community acquired pneumonias. They demonstrated an area under the curve (AUC) of 0.959 with a sensitivity and specificity of 0.879 and 0.887 respectively 24 . Gangloff et al, evaluated machine learning models including logistic regression, random forest and neural networks, using routine clinical and laboratory data to improve the performance of RT-PCR and chest-CT for COVID-19 diagnosis.…”

Transcriptome-Based Molecular Networks Uncovered Interplay Between Druggable Genes of CD8+ T Cells and Changes in Immune Cell Landscape in Patients With Pulmonary Tuberculosis

“…They perform variety of biological analyses including immune responses based on the defined gene sets. Exploring abnormal immune cell infiltration is critical for developing novel transformative therapies to combat diseases such as cancer, myocarditis, and TB (11,12). Therefore, in order to characterize alterations in immune cell proportion landscape and transcriptomic profile, and to identify new molecular therapeutic targets, this study applied statistical and knowledgebased systemic investigations (such as semantic similarity, gene correlation, and graph theory parameters) to the blood transcription data of patients with TB.…”