infer-cnv-visualization

Visualization is heavily influenced by https://github.com/ColinEberhardt/d3fc-webgl-hathi-explorer

To produce the visualization there are two main steps:

1. Data preparation
2. Serving the website

Data Preparation

There are three files you need.

The first file is infercnv.heatmap_thresholds.txt which is found inside of your infercnv output

The second and third files are spot.tsv and output_compressed.tsv. These files are produced by the data_prep directory.

The conda environment is located at: /diskmnt/Projects/Users/efang/miniconda3/envs/FindNormal2

Once the environment is built, run conda activate FindNormal2

cd into the data_prep directory and run: python main.py {spaceranger_output_dir} {annotation_file} {infercnv.observations.txt_file_located_in_infercnv_output} {gene_order_file}

An example of this run is: python main.py /diskmnt/Datasets/Spatial_Transcriptomics/outputs_OCT/Human/HT206B1/H1/HT206B1-U1_ST_Bn1/outs /diskmnt/Projects/Users/efang/normalized_annotations2/HT206B1-U1_ST_Bn1-with_normal_annotations.csv /diskmnt/Projects/Users/efang/SpatialInferCNV/InferCNVRunsOutput2/InferCNVrun_outputs-HT206B1-U1_ST_Bn1/infercnv.observations.txt /diskmnt/Projects/Users/efang/SpatialInferCNV/siCNV_GeneOrderFile.tsv

This will output both spot.tsv and output_compressed.tsv in the same directory (as well as some byproduct files)

Move all three files into the same directory that you will serve the website with.

Serving the website

The easiest way to serve the website is using VSCode.

You can also serve the website through Python flask if you want.

Todo:

• Create a legend for the tumor annotations
• Modify the pointer for the bottom image so that it can also create annotations to the top