Cloud-based single-cell copy-number variation analysis tool
PHP JavaScript R Shell C++ CSS Other
Switch branches/tags
Nothing to show
Clone or download
Michael Schatz
Michael Schatz add
Latest commit 9ad3be4 Jan 3, 2018
Permalink
Failed to load latest commit information.
cli CLI - Cromwell: put ginkgo logic into WDL Oct 31, 2016
genomes/scripts Patch for issue #6 May 11, 2017
includes refactor, reset default paths for new qb.cshl.edu server Mar 3, 2017
scripts add Jan 3, 2018
LICENSE Add license Feb 10, 2016
Makefile Add CNVcaller to the list of software to be compiled Feb 25, 2016
README.md Update README.md Mar 3, 2017
README.standalone Refactor analyze Jun 28, 2016
admin.sh Refactor analyze Jun 28, 2016
bootstrap.php
config.example Include instructions to run GINKGO on the command line Apr 4, 2016
index.php
loading.gif Added loading... image Jun 21, 2014

README.md

Ginkgo

Ginkgo is a cloud-based single-cell copy-number variation analysis tool.

Launch Ginkgo: qb.cshl.edu/ginkgo

Usage

  • Step 0: Upload .bed files
  • Step 1: Choose analysis parameters
  • Step 2: Compute Copy Number Profiles, and a Phylogenetic Tree
  • Step 3: Analyze Individual Cells

Setup Ginkgo on your own server

Requirements:

  • PHP >=5.2
  • R >= 3.0.0
  • R Packages:
    • ctc
    • DNAcopy
    • inline
    • gplots
    • scales
    • plyr
    • ggplot2
    • gridExtra
    • fastcluster
    • heatmap3

WARNING Version 3.0.0 (Mar 28, 2016) of gplots introduced a bug in heatmap.2 that makes it calculate dendrograms even when Rowv or Colv is set to FALSE so that Ginkgo will run for a very long time. The solution is to use an older version of gplots or use fixed version from: https://github.com/ChristophH/gplots. This can be installed using the commands below

remove.packages('gplots'); 
library('devtools'); 
install_github("ChristophH/gplots")

Install Ginkgo:

Type make in the ginkgo/ directory

Server Configuration:

  • /etc/php.ini

    • upload_tmp_dir: make sure this directory has write permission
    • upload_max_filesize: set to >2G since .bam files can be large
  • ginkgo/includes/fileupload/server/php/UploadHandler.php

    • In constructor, on line 43 and 44:
      • upload_dir = [FULL_PATH_TO_UPLOADS_DIR] . $_SESSION["user_id"] . '/'
      • upload_url = [FULL_URL_TO_UPLOADS_DIR] . $_SESSION["user_id"] . '/'
  • ginkgo/bootstrap.php

    • Change DIR_ROOT, DIR_UPLOADS and URL_ROOT
  • ginkgo/scripts/analyze.sh

    • Change home variable to where the ginkgo/ folder is located
  • ginkgo/scripts/process.R

    • Change main_dir variable to the folder where ginkgo/scripts is located
  • ginkgo/scripts/reclust.R

    • Change main_dir variable to the folder where ginkgo/scripts is located
  • ginkgo/scripts/analyze-subset.R

    • Set the folder to where ginkgo/scripts is located
    • Set the folder to where ginkgo/genomes is located (warning: test this carefully if your ginkgo/uploads folder is a symlink)
  • Make sure the uploads directory has the correct write permissions

Download data files: