CST Lung Cancer Viz

This repo contains

• Lung cancer cell line PTM and gene expression data from CST and CCLE, respectively
• Data-Processing Python scripts
• IPython/Jupyter notebooks describing processing/normalization
• the source code for the page CST_Lung_Cancer_Viz

Lung Cancer Data

Our collaborators at Cell Signaling Technology (CST) used SILAC mass spectrometry to measure differential phosphorylation, acetylation, and methylation in a panel of 42 lung cancer cell lines compared to non-cancerous lung tissue (the primary data is here). Gene expression data from 37 of these lung cancer cell lines was obtained from the publically available Cancer Cell Line Encyclopedia (CCLE).

CST PTM Data Overview

The PTM data from CST is overviewed in two stages: 1) overview of the data, 2) overview of the normalization procedure

Data and Missing-Data Overview

The notebook, CST_PTM_Data_Overview.ipynb, overviews the PTM data and discusses the reasoning behind our method of normanlization and filtering.

Normalization Overview

The notebook, CST_PTM_Normalization_Overiew.ipynb

The notebook, t-SNE_Cell_Line_Clustering_Phosphorylation.ipynb, uses the dimensionality reduction algorithm, t-SNE, to visualize cell line clustering based on phosphorylation data before and after normalization and filtering.

CCLE Gene Expression Data Overview

The python notebook, CCLE_Gene_Expression_Data_Overview, overviews the gene expression data and discusses some results from our analysis of this data.

The notebook, t-SNE_Cell_Line_Clustering_Gene_Expression.ipynb, uses t-SNE to visualize cell line clustering based on gene expression data.

Data Processing Scripts

All data was processed using Python scripts in two broad steps: 1) data was pre-processed (e.g. calculating ratios of cancer vs non-cancer levels) and combined into a simple tab-separated format and 2) data was normalized and filtered in order to make heatmap visualizations. Visualizations for the webpage were made using the Clustergrammer web-based visualization tool. The clustergrammer python module was used to normalize/filter data and produce JSONs for clustergramer.js .

Data Pre-processing

The script process_latest_cst_data.py was used to 1) calculate the log2 peak ratios of lung cancer cell lines to their associated Normal Pool from the same plex, and 2) combine these ratios into a single tab-separated file.

Data Normalization and Filtering

The script make_cst_homepage_figures.py was used to make Clustergrammer interactive visualizations for the website. Specifically, this script normalizes and processes the data and creates the JSONs for the front-end visualizations.

Webpage Source Code

This repo contains the souce code for the site CST_Lung_Cancer_Viz. The page can also be seen through github.io https://maayanlab.github.io/CST_Lung_Cancer_Viz/ with some limited capabilities (due to github.io HTTPS requirements).