Genomic and Transcriptomic Analysis for Primary and Multi-metastatic Tumor Samples

We anaylzed deep whole genome sequencing data and RNA sequencing data of 30 tumors from a metastatic breast cancer patient. Here are the scripts that we used for analyses. Results are published in manuscript "Novel temporal and spatial patterns of metastatic colonization from rapid-autopsy tumor biopsies" (https://doi.org/10.1101/2021.06.03.446803).

We divided the analyses to the following 5 sections:

1. CNV analysis from WGS
2. Phylogeny inference
3. Subclone reconstruction
4. Mutational Signature analysis
5. RNAseq analysis

In each section, we provide necessary input files (except controlled access bam files) and scripts to perform the analysis.

CNV analysis from WGS

See the README.md file in the CNV analysis folder

Phylogeny inference

We used two methods to study the phylogenetic relationships among tumor samples, which were described in the manuscript. We provide phylogeny.R script and an input vcf file to generate an allele frequency heatmap for all samples, as well as to perform UPGMA clustering based on hamming distance. We also provide seederSeeker.py script and an input txt file (contains allele frequency values) to study the phylogenetic relationships among tumor samples based on the perfect phylogeny algorithm. There are two parameters in the script: one is a threshold for allele frequency in a sample to determine whether a variant is present or absent; the other one is a cutoff for the minimun number of variants that have the same sample distribution.

The script can be invoked in the following manner

$ cat strictSomatic_correctAF.txt | python3 seederSeeker.py > output.dot

The output file can be visualized by graphvis

Subclone Reconstruction

We used SubcloneSeeker V2 to generate subclone structures. We refer you to the SubcloneSeeker repository for more information. We provide input files, and a Makefile to perform all analysis once SubcloneSeeker is installed

Mutational Signature

We provide ra_mutationpatterns.R script to perform mutational signature analysis. We provide input vcf files for the above R script.

RNAseq analysis

We provide CNV_RNAseq.R script to perform copy number inference from RNAseq data. In this R script, we provide the function for PCA and clustering analysis.

We provide edgeR_analysis.R for differentially expressed gene analysis.

The input files required for above analysis are included.