#model #genomic #variant #variable #variants #free #parameter

bin+lib varlociraptor

A library for calling of genomic variants using a latent variable model

41 stable releases

new 3.1.0 May 4, 2021
2.6.5 Mar 19, 2021
2.5.3 Nov 23, 2020
2.0.0 Jul 9, 2020
1.1.1 Jul 9, 2019

#63 in Science

Download history 39/week @ 2021-01-14 9/week @ 2021-01-21 9/week @ 2021-02-04 117/week @ 2021-02-11 72/week @ 2021-02-18 2/week @ 2021-02-25 18/week @ 2021-03-04 42/week @ 2021-03-18 153/week @ 2021-03-25 38/week @ 2021-04-01 51/week @ 2021-04-08 121/week @ 2021-04-15 63/week @ 2021-04-22 63/week @ 2021-04-29

244 downloads per month

GPL-3.0 license

605KB
14K SLoC

Varlociraptor

Bioconda GitHub Workflow Status Codecov API docs Commitizen friendly

Varlociraptor implements a novel, unified fully uncertainty-aware approach to genomic variant calling in arbitrary scenarios.

Key features

  • Calls SNVs, MNVs, indels, inversions, duplications, replacements and breakends in all length ranges (from small to structural) with a unified statistical model.
  • The statistical model encompasses all possible sources of uncertainty and biases.
  • Resulting variant calls can be filtered by false discovery rate. No parameter tuning necessary.
  • Unbiased, maximum a posteriori allele frequency estimates are provided with each call.

Calling modes

  • Tumor-normal-calling, classifying variants as somatic in tumor, somatic in normal, germline, or absent.
  • Generic, grammar based configuration of the statistical model, allowing to classify arbitrary scenarios, from poplation genetics, to pedigrees, complex tumor scenarios and arbitrary combinations thereof.

For details, see the homepage: https://varlociraptor.github.io

Dependencies

~41MB
~1M SLoC