This is a workflow for running CLUES on the computationally phased 1000 genomes data. There are two workflow files: workflow_1000g_derived_info.py and workflow_clues.py. The workflow_1000g_derived_info.py workflow needs to be run first to generate a HDF5 file that is used by the main workflow workflow_clues.py to look up information about derived variants. Both workflows hardcoded paths to data files in /project/simons/faststorage/data/1000Genomes. So membership of the simons project folder is required.
It is assumed that clues-v0 is installed (unpacked) in the base directory of the repository.
You need to make a conda envirommnet, E.g. clues-v0, with at least these packages:
conda create -n <some name> -c anaconda -c conda-forge -c genomedk -c gwforg -c kaspermunch -c bioconda vcftools gwf pandas numpy h5py biopython future argweaver scikit-allel chromwindow
Activate environemnt and run like this:
gwf -f workflow_1000g_derived_info.py run
Once it finishes you run:
gwf -f workflow_clues.py
This workflow produces a table file with LR scores for all analysed SNPs.
The notebooks folder contains a notebook with hints to the interpretation of raw CLUES output.