Pipeline for the analysis of synthetic lethal screens in human haploid cells

1. Installation

1. Either clone the GitHub repository or extract the downloaded zip archive at your favorite place
2. Ensure yourself that you have installed the following packages
   • Python 3
   • ktinker (apt-get install python3-tk or yum install python34-tkinter)
   • statsmodels
   • pandas
   • scpipy
   • numpy
3. Optional step if you want the required Python libraries to be installed in a virtual environment rather than systemwide
   virtualenv -p python3 venv
source venv/bin/activate
4. Install the following packages using pip
   pip install matplotlib numpy pandas

Preparing the control dataset

• If you have installated the packages in a virtual environment, open the virtual environment by typing
  source venv/bin/activate
• For each of the four controls run analyze_sli.sh using the following options
  ./analyze_sli.sh -C -R=<1-4> -S=sequence_reads.fastq.gz --name=ControlData-HAP1
• Update the analytical_settings.py configuration file and make sure each 'HAP1_control' points to the corresponding sense_vs_antisense_counts.csv file that now resides in the final directory that is listed in settings.conf
• Now run the following command to normalize the counts of the controls:
  python3 sub/normalize.py --normalize=controls --outdir=<some directory where you want to store the normalized controls> --screenname=<screenname> --refname=<name of the reference annotation, same as entered in alignment_settings.conf>
  This will create two files for each replicate:
  1. ControlData-HAP1_replicate_1_sense_vs_antisense_counts_normalized.csv
  2. ControlData-HAP1_replicate_1_for_phenosaurus.csv
  • Update the analytical_settings.py configuration file and make sure each 'HAP1_control' points to the corresponding sense_vs_antisense_normalized_counts.csv file that now resides in the temp directory.
• You are now ready to analyze your first synthetic lethal screen