Bimodal gene expression in cancer patients provides interpretable biomarkers for drug sensitivity

The Hallmarks of Predictive Oncology

SIGNIFICANCE

As the field of artificial intelligence evolves rapidly, these hallmarks are intended to capture fundamental, complementary concepts necessary for the progress and timely adoption of predictive modeling in precision oncology. Through these hallmarks, we hope to establish standards and guidelines that enable the symbiotic development of artificial intelligence and precision oncology.

Text-mining-based feature selection for anticancer drug response prediction

Motivation

Predicting anticancer treatment response from baseline genomic data is a critical obstacle in personalized medicine. Machine learning methods are commonly used for predicting drug response from gene expression data. In the process of constructing these machine learning models, one of the most significant challenges is identifying appropriate features among a massive number of genes.

Results

In this study, we utilize features (genes) extracted using the text-mining of scientific literatures. Using two independent cancer pharmacogenomic datasets, we demonstrate that text-mining-based features outperform traditional feature selection techniques in machine learning tasks. In addition, our analysis reveals that text-mining feature-based machine learning models trained on in vitro data also perform well when predicting the response of in vivo cancer models. Our results demonstrate that text-mining-based feature selection is an easy to implement approach that is suitable for building machine learning models for anticancer drug response prediction.

Availability and implementation

https://github.com/merlab/text_features.

Prognostic and immune infiltrative biomarkers of CENPO in pan-cancer and its relationship with lung adenocarcinoma cell proliferation and metastasis

BACKGROUND

The centromere protein O (CENPO) is an important member of the centromere protein family. However, the role of CENPO in pan-cancer and immune infiltration has not been reported. Here, we investigated the role of CENPO in pan-cancer and further validated its role in lung adenocarcinoma (LUAD) by in vitro experiments.

METHOD

The UCSC Xena database and The Cancer Genome Atlas (TCGA)-LUAD data were used to assess the expression levels of CENPO. The potential value of CENPO as a diagnostic and prognostic biomarker for pan-cancer was evaluated using TCGA data and the GEPIA database. The -expression profiles of LUAD patients and the corresponding clinical data were downloaded for correlation analysis. The role of CENPO in immune infiltration was investigated using the UCSC Xena database. Subsequently, qRT-PCR was performed to detect the expression of CENPO. Cell proliferation, migration, and invasion were determined using CCK-8, wound-healing assay, and transwell assay, respectively.

RESULTS

CENPO is highly expressed in most cancers, and the upregulation of CENPO is associated with poor prognosis in many cancers. CENPO expression correlates with age, TNM stage, N stage, T stage, and receipt of radiotherapy in LUAD patients, and LUAD patients with high CENPO expression have poorer overall survival (OS) and disease-free survival (DFS). In addition, CENPO expression is associated with immune cell infiltration and immune checkpoint inhibitors. Moreover, the expression of CENPO was closely related to the expression of tumor mutational load and microsatellite instability. In vitro experiments showed that CENPO expression was increased in LUAD cell lines and that knockdown of CENPO significantly inhibited the proliferation, cell invasion, and migration ability of LUAD cells.

CONCLUSION

CENPO may be a potential pan-cancer biomarker and oncogene, especially in LUAD. In addition, CENPO is associated with immune cell infiltration and may serve as a new molecular therapeutic target and effective prognostic marker for LUAD.

Neurocan expression associates with better survival and viral positivity in Merkel cell carcinoma

Merkel cell carcinoma (MCC) is a rare cutaneous neuroendocrine carcinoma that is frequently divided into Merkel cell polyomavirus negative and positive tumors due their distinct genomic and transcriptomic profiles, and disease outcomes. Although some prognostic factors in MCC are known, tumorigenic pathways, which that explain outcome differences in MCC are not fully understood. We investigated transcriptomes of 110 tissue samples of a formalin-fixed, paraffin-embedded MCC series by RNA sequencing to identify genes showing a bimodal expression pattern and predicting outcome in cancer and that potentially could play a role in tumorigenesis. We discovered 19 genes among which IGHM, IGKC, NCAN, OTOF, and USH2A were associated also with overall survival (all p-values < 0.05). From these genes, NCAN (neurocan) expression was detected in all 144 MCC samples by immunohistochemistry. Increased NCAN expression was associated with presence of Merkel cell polyomavirus DNA (p = 0.001) and viral large T antigen expression in tumor tissue (p = 0.004) and with improved MCC-specific survival (p = 0.027) and overall survival (p = 0.034). We conclude that NCAN expression is common in MCC, and further studies are warranted to investigate its role in MCC tumorigenesis.