primal - orenlivne/ober GitHub Wiki

PRIMAL (PedigRee-based IMputation Algorithm) is an accurate phasing and imputation algorithm for related individuals. It is developed by the Ober Lab, Department of Human Genetics, The University of Chicago. PRIMAL was originally written to impute the genomes of 1,317 members of the South Dakota Hutterites population from 98 whole-genome Hutterite sequences. The software is written in python and can be run in three different modes:

  • Programmatic python calls (API). This allows you to write PRIMAL wrappers for your own application with little effort.
  • Serial mode. Just run our programs at the command line on a single-processor machine.
  • Parallel pipelines for a cluster with a Torque/PBS scheduler, if you have a large data set.

See Setup instructions for setting up the $OBER directory and the rest of the environment.

Main Steps

  1. Phasing
  2. IBD segment identification
  3. IBD segment indexing
  4. Imputation
  5. Parental origin assignment
  6. IMPUTE2 post-processing (to increase the call rate)

Parallel Pipelines

PRIMAL pedigree-based imputation

Main script: $OBER/impute/batch/cgi/pipeline-impute-cgi

LD-based post-processing imputation

Main script: $OBER/impute/batch/cgi/pipeline-impute2

  • Requires installing impute2.

LD-based post-processing parental origin assigment

Main Script: $OBER/impute/batch/impute2/pipeline-impute2-po

Add a custom footer
⚠️ **GitHub.com Fallback** ⚠️