doi:10.5281/zenodo.5542001


For training, go to bob/train folder and do

make clean-all
make all
make deploy

Above will create the training data when the tag SNPs are 10k bp apart from each other (using sorted_tag_SNPs_10k_genotypes.data) for 500 target SNPs in chromosome 1 (sorted_target_SNP_genotypes.data)

Then go to the main folder


To compile:
make compile

To run:
make run ID=id_of_database QUERY=name_of_the_query_file
e.g.:
make run ID=1 QUERY=query_tag_SNPs_1_genotypes.data [for this, the training needs to be done sorted_tag_SNPs_10k_genotypes.data, whihc are tag SNPs 1k apart from each other]
make run ID=10 QUERY=query_tag_SNPs_10_genotypes.data

The outputs are:
- ypredID.data has the probabilities
- targetID.data has the predicted values
- timeID.data has the execution time (round trip, encryption, computation, decryption)

If you prefer to use the Docker file, please create a folder called idash and upload everything but Dockerfile and run_idash.sh in that folder.