Integrated bioinformatics for oncology research on solid tumors and the tumor microenvironment: a proposed model for in silico studies

Keywords

transcriptome - bioinformatics - oncology

Introduction: Since the Human Genome Project, the volume of Omics data has required complex in silico analyses strategies, as the data volume exceeds human analytical capacity. The literature reflects a growing number of genomic and transcriptomic information derived from robust databases, guiding clinically relevant hypotheses. However, there is no standardization or defined methodological strategies, complicating the interpretation, validation and comparison of results from different groups.

Objective: To propose an integrated in silico workflow using open databases, aiming to standardize the analysis of the transcriptome of solid tumors in a translational research laboratory linked to graduate programs in Human Pathology and Oncology.

Methods: An integrated workflow was established for transcriptome analysis using The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) databases. Data analysis algorithms were selected via the collaborative platform GitHub, and the R environment was employed for the in silico study.

Results: Guided by clinical questions, two experienced oncologists selected and curated the available databases in TCGA and GEO. Transcriptome data (RNASeq) were downloaded into the R environment, followed by the identification of differentially expressed genes (DEG), using the Limma and/or DESeq2 packages, subsequently grouped into weighted gene co-expression networks (WGCNA and GeneXPress). Functions related to gene communities were obtained through gene ontology on the DAVID, KEGG, and Medscape platforms. The integrated analysis between the gene signature and aspects of the tumor microenvironment was conducted on the CIBERSORT and TIMER platforms, which provide relevant data on immune cells, tumor and non-tumor gene expression. Additionally, the study of predicted protein interaction networks was conducted using the String platform, identifying the most relevant proteins within each previously described community of interest.

Conclusion: The establishment of a coherent omics data analysis workflow allows for the standardization of transcriptome and tumor microenvironment studies across various solid tumors, utilizing information obtained from cancer databases. This has significant implications for the reproducibility, consistency, and applicability of results, promoting transparency and generating new hypotheses that guide complementary studies, thereby expanding relevant scientific knowledge for clinical practice.

Corresponding author: Priscila Doria (e-mail: [e-mail]).

 

Scroll to top

No conflict of interest has been declared by the author(s).

Publication History

Article published online:
06 May 2025