Sovereign 8s.is mirror of standardgalactic/delve
Find a file
Repository files (latest commit first)
Filename Latest commit message Latest commit date
Bernd Bohmeier 884a31ccfd
Some checks failed
CI / build (push) Has been cancelled
CI / lint (push) Has been cancelled
Add acknowledgements for funding sources
Added acknowledgements section for funding sources.
2025-11-13 01:40:32 +01:00
.github/workflows Add linting and fmt check 2025-09-03 19:28:57 +02:00
benches Initial commit, variant caller 2025-08-14 16:40:57 +02:00
src Remove truncate region feature 2025-09-29 18:19:17 +02:00
.gitignore Initial commit 2025-08-14 16:25:16 +02:00
Cargo.lock Add bett error handling 2025-09-03 19:11:11 +02:00
Cargo.toml Add bett error handling 2025-09-03 19:11:11 +02:00
LICENSE Initial commit 2025-08-14 16:25:16 +02:00
README.md Add acknowledgements for funding sources 2025-11-13 01:40:32 +01:00

CI Conda Downloads OS - Linux | OSX install with bioconda

delve

Delve is a SNP variant caller for mixed infections, like they occur in Malaria. We primarly use it to call SNPs in plasmodium falciparum. It is able to call SNPs of minor clones at low percentages.

Install

Delve can be install from bioconda

conda install bioconda::delve-bio

Run

The minimum you need to run delve is an indexed BAM file with aligned reads to a reference genome, for example with minimap2, a reference genome in fasta format with index. Optionally you can provide a BED file with regions over which to call the variants.

delve -f reference.fasta [-R amplicons.bed] input.bam

Features

Delve internally uses rust-htslib to build a pileup of reads over each position, then uses a statistical model, which uses the base read errors of the reference and alternative bases on this position to calcululate a likelihood of a proportion p of alt vs. ref bases, and then uses a Likelihood-ratio Test to determine if there is enough evidence that a variant exists. After, it uses a strand bias test to filtered out wrong calls. Currently delve only calls variants on one sample at a time.

New Features

We plan to implement a cluster based approach in the future to be able to call (micro) haplotypes. If you have other needs, like calling at multiple samples at a time, open an issue, and I might implement it.

Contribute

Contributions and suggestions are welcome. Feel free to open an issue with suggestions and we can talk about new features and potential pull requests, or email me at bohmeier[at].mpiib-berlin.mpg[dot]de

Acknowledgements

This work was funded by the Bill and Melinda Gates Foundation (INV-003660, INV-048316).