- Download an
archive containing the GImap scripts and
example data. Unzipping of the archive will create
the required subdirectory structure.
- Download and install bowtie.
- Install python, or, in order to use interactive graphics, install the Canopy Python package, which is free for academic use.
- To generate tab-delimited text files, we recommend the use of a command-line text editor, like vim.
- Download Homo_sapiens.gene_info.gz from the NCBI, unzip the file and move it to subdirectory LUT.
- Download and install the Pycluster module.
- Download and install Cluster 3.0.
- Download and install Java Tree View.
- Create a
fasta file for the shRNA library you used for the primary screen as
follows: Each entry describes one shRNA. Each shRNA name should begin
with the Entrez Gene ID, followed by a double underscore, and a number
or other information that uniquely identifies the shRNA, e.g.
9064__NM_004672.3__24. Names of negative control shRNAs should have a 0
instead of an Entrez Gene ID, e.g. 0__0__1. The nucleotide sequence
included for each shRNA is the reverse-complement of the shRNA guide
strand for the shRNA in your shRNA library. (The length of each
sequence should be 22 nt, so in case your guide strand is shorter,
include the reverse-complement sequence of the constant nucleotides
following the guide strand in your shRNA construct.) To create a bowtie
index from this file, use the bowtie-build command. For example, if the
path to bowtie-build on your computer is
/Applications/bowtie-0.12.3/bowtie-build, and the name of the fasta
file for your shRNA library is LibraryName.fa, enter the following on
the command line:
/Applications/bowtie-0.12.3/bowtie-build LibraryName.fa LibraryName
Six output files are created, move them to the subdirectory indices/.
In the script align_primary.py (subdirectory scripts), specify the path
to bowtie on your computer in line 4.