treeclimbR_article

This repository provides toy examples to understand treeclimbR and codes to reproduce figures in treeclimbR article. The used R version was 3.6.1 (2019-07-05)

Installation

• Data container: TreeSummarizedExperiment
• Algorithm: treeclimbR
• Data visualization: TreeHeatmap.

# TreeSummarizedExperiment: version (1.3.1)
remotes::install_github(“fionarhuang/TreeSummarizedExperiment@dfa2097c61f900c3b2989a4f045f864b992fc8c5")
# treeclimbR: version (0.1.1)
remotes::install_github("fionarhuang/treeclimbR@treeclimbR_article")
BiocManager::install("fionarhuang/TreeHeatmap_old")

Toy examples (click here)

Below are results of one toy dataset.

• To capture signal patterns on the tree, treeclimbR firstly propose multiple candidates that are generated by tuning a parameter t:

Heatmap shows counts of entities (rows) in samples (columns) split by groups. Branches that include entities with signals are colored in orange.

• Nodes detected by from treeclimbR are compared to those detected by BH under FDR 0.05.

Reproduce figures in the manuscript of treeclimbR

1. Parametric synthetical microbial data (in folder simulation_microbe)

• install snakemake & conda
• Clone this repository and set directory to simulation_microbe/
• Modify the configuration file (config.yaml) to specify paths of input and output files.
  • Rbin: the path to the system R
  • condaR: the path to conda R
    • run create an environment: conda env create -f envs/lefse.yml
    • run conda activate lefse
    • run which R to get path to conda R
• Specify the R library paths in the .Renviron. If there is less than 3 library paths, then set R_LIBS_3 = "" to remove the third library path.
• dry run the pipeline using snakemake --use-conda -npr
• run the pipeline using snakemake --use-conda --cores n (n is the number of cores to be used)
• Once the pipeline is run successfully, all figures could be generated using all_figure.R under the folder simulation_microbe/summary/.

2. Non-parametric synthetical microbial data (see here)

3. AML-sim and BCR-XL-sim (in folder cytof)

The semi-simulated data should be downloaded in the cytof/data/ folder from HDCytoData using Download.R before running the pipeline

• AML-sim (DA folder)

  • install snakemake
  • Set directory to DA/
  • Specify paths to input and output files in the configuration file (config.yaml)
  • Specify R library paths in the .Renviron. If there is less than 3 library paths, then set R_LIBS_3 = "" to remove the third library path.
  • dry run the pipeline using snakemake -npr
  • run the pipeline using snakemake --cores n (n is the number of cores to be used)

• BCR-XL-sim (DS folder). Similarly to run AML-sim

Figues 3 is generated using all_figure.R

4. Infant gut microbial data (in folder microbe)

5. Mouse miRNA data (in folder miRNA)

6. Mouse cortex scRNAseq data(in folder LPS)