The Lung-Tumor Microenvironment Interactome

The Lung-Tumor Microenvironment Interactome application is based on analysis from Gentles et al. (20XX). We performed RNA-seq profiling of flow-sorted malignant cells, endothelial cells, immune cells, and fibroblasts from human primary non-small-cell lung tumors (Lung Adenocarcinoma, Lung Squamous Cell Carcinoma). We mapped the cell-specific differential expression of prognostically-associated secreted factors and cell surface genes, and computationally reconstructed cross-talk between these cell types. This application allows users to complete the same analyses on custom genes and cell types. The following tutorial reproduces one of the analyses from Gentles et al.

1. Select a gene of interest

Select GREM1 by searching for it in the sidebar. Observe its characteristics and expression in different cell types.

Not all genes are initially displayed in the dropdown menu. To select your gene of interest, search for its prefix and it will be displayed as an option.

To filter the list of displayed genes based on prognistic, surfacome, and expression data, select criteria using the checkboxes.

Notably, GREM1 is prognostic in Adenocarcinoma, expressed on the surfaceome, and is expressed highly on one cell type (Fibroblasts).

2. Select a cell type

Select Fibroblasts to calculate the pearson correlations between GREM1 expression in fibroblasts and every other gene/celltype pair.

Observe the heatmap of expression correlations and sort by a cell type to find the genes that are most correlated with GREM1 in fibroblasts.

Click on a heatmap cell to view a scatterplot of the expressions and significance of the correlation.

To download a .tsv file of all correlations for the gene of interest and cell type, click the 'download tsv' button.

3. Run GSEA analysis

Select a comparison celltype for GSEA (Malignant)

We will select the malignant population to determine which gene sets and biological pathways are enriched in malignant cells when GREM1 is overexpressed in fibroblasts.

Select one of the MSIGDB gene set collections. We select the msigdb c2 + c5 bioprocess category as this was used in Gentles et al.

Navigate to the GSEA tab, and wait for the analysis to complete

Observe the significantly enriched gene sets in adeno and squamous carcinoma

To download the full GSEA results, click the 'download tsv' button

Data can be downloaded at GEO Accession: GSE111907

Access PRECOG Data Here

For questions about the analysis performed on this site, contact:

Andrew Gentles

Assistant Professor (Research) of Medicine (Biomedical Informatics) and, by courtesy, of Biomedical Data Science

Stanford University School of Medicine, Stanford CA 94305

andrewg@stanford.edu

For questions about the publication associated with this site, contact:

Sylvia K. Plevritis

Professor of Biomedical Data Science and of Radiology

Stanford University School of Medicine, Stanford CA 94305

sylvia.plevritis@stanford.edu

Max Diehn

Associate Professor of Radiation Oncology (Radiation Therapy)

Stanford University School of Medicine, Stanford CA 94305

diehn@stanford.edu

For questions about the code behind this site, contact:

Armon Azizi