bmibdcaller_empirical

This repository contains code used for validation of IBD caller benchmarking and optimization findings in emprical data sets.

The pipeline is part of a broader project bmibdcaller, which also includes the following repositories:

• bmibdcaller_simulations for simulations analysis.
• ishare/ibdutils for effienciently comparing two sets of inferred IBD segments.
• tskibd for obtaining true ibd segments from simulated genealogical trees.

There are two main components in the repository, including

1. Scripts and notes to pre-process data available from MalariaGEN Pf7.
   • Data pre-filtering and downloading, inferring dominant alleles, and imputation. See notes ./input/Readme.md.
   • Constructing different datasets. See notes ./datasets/Readme.md.
2. The Nextflow pipeline to benchmark the performance of multiple IBD callers before and after IBD caller parameter optimization:
   • Pipeline entry.

For the emprical analysis pipeline, the software environment and result folder structure of this pipeline are similiar to the bmibdcaller_simulations repository. Details can be found Follow the readme from the simulation repository.

Examples of running the pipeline can be found in

• runs/r240113_opt/run.sh.
• runs/r240114_noopti/run.sh.