Skip to content

lohhlamod

Introduction

lohhlamod is the original LOHHLA algorithm re-engineered in modern style that

  • offers additional features/metrics for better interpretation
  • offers runtime speedup
  • separates HLA realignment from LOH detection
  • maximizes vectorized operations whenever possible
  • provides better output layout and makes intermediate result per allele in RDS format accessible
  • provides proper packaging for ease use in multi-user HPC environment
  • removes hardcoded path presets for better code maintenance

Installation

It is recommended to use docker for building and running lohhlamod.

docker compose build --pull

Add --no-cache option for a refresh build.

Verify installation

After installation, I recommend first running lohhlamod using the sample simulation data provided in the repository, and verify that your output matches the expected results. The expected results can be found under the output/ subdirectory within each simulated case. Refer to the Dataset for details about simulation.

Citation

Please cite the original HLALOH paper, its Bitbucket repository, and the lohhlamod repository.

License

  • lohhlamod fully respects all LICENSE requirments imposed by the original LOHHLA tool.
  • lohhlamod is free to use for all non-commercial parties.

Disclaimer

  • I, by no means, intent to overtake the origianl idea, implementation, and copyright of the original LOHHLA algorithm.
  • This repo does not distribute LOHHLA package, as well as all its dependent third-party parties that are under commercial licenses.
  • lohhlamod does not necessarily produce identical result as LOHHLA, including but not limited to estimates of copy numbers, logR, BAF, plots, etc.
  • Please interpret result at your own discretion when using lohhlamod. I simulated a few datasets to demonstrate the performance of lohhlamod in the most ideal settings.