All Projects → SamStudio8 → gretel

SamStudio8 / gretel

Licence: MIT license
An algorithm for recovering haplotypes from metagenomes

Programming Languages

python
139335 projects - #7 most used programming language
Makefile
30231 projects

Projects that are alternatives of or similar to gretel

starfish
Intersect multiple VCF files with haplotype awareness
Stars: ✭ 19 (-36.67%)
Mutual labels:  haplotypes
minorseq
Minor Variant Calling and Phasing Tools
Stars: ✭ 15 (-50%)
Mutual labels:  haplotypes
CliqueSNV
No description or website provided.
Stars: ✭ 13 (-56.67%)
Mutual labels:  haplotypes
PlasFlow
Software for prediction of plasmid sequences in metagenomic assemblies
Stars: ✭ 74 (+146.67%)
Mutual labels:  metagenomes
pyrodigal
Cython bindings and Python interface to Prodigal, an ORF finder for genomes and metagenomes. Now with SIMD!
Stars: ✭ 38 (+26.67%)
Mutual labels:  metagenomes

gretel-logo

Gretel

An algorithm for recovering haplotypes from metagenomes. Sister to Hansel.

License bioconda

What is it?

Gretel is a Python package providing a command line tool for the recovery of haplotypes from metagenomic data sets. Gretel parses an alignment of reads into a Hansel matrix and uses the evidence of SNP pairs observed to appear on the same reads to probabilistically reconstruct the most likely haplotypes.

Gretel uses an L'th order Markov chain model to reconstruct likely sequences of variants that constitute haplotypes in the real metagenome. Our approach involves graph-like traversal of the data within the Hansel matrix. Edges are probabilitically weighted based on the evidence on the reads, as well as the haplotype as it has been reconstructed so far.

What can I use it for?

Gretel is designed to recover haplotypes from your data set, without the need for setting (or optimisation) of any parameters. Gretel does not require a priori knowledge of your input data (such as its contents, or the true number of haplotypes) and makes no assumptions regarding the distributions of alleles at variant sites and uses the available evidence from the aligned reads without altering or discarding the observed varations.

Why should I use it?

Gretel is the first tool capable of recovering haplotypes from metagenomes. Whilst tools exist for analogous haplotyping problems, such as the assembly of viral quasispecies, typically these tools rely on overlap approaches that create too many unranked haplotypes. Gretel is capable of ranking the haplotypes it outputs by their likelihood.

Gretel requires no parameters and our approach is robust to sequencing error and misalignment noise.

Requirements

$ pip install numpy hanselx pysam PyVCF

Install

$ pip install gretel

Alternatively, Gretel has been packaged for bioconda (Thanks @johnne!):

$ conda install -c bioconda gretel

Usage

You will require a sorted BAM containing your reads, aligned to some pseudo-reference. You can use any sequence as your reference, such as a consensus assembly of the metagenomic reads, or a known strain reference (such as HIV-1). You must bgzip and tabix your VCF.

$ gretel <bam> <vcf.gz> <contig> -s <1-start> -e <1-end> --master <master.fa> -o <outdir>

Citation

@article{10.1093/bioinformatics/btaa977,
    author = {Nicholls, Samuel M and Aubrey, Wayne and De Grave, Kurt and Schietgat, Leander and Creevey, Christopher J and Clare, Amanda},
    title = "{On the complexity of haplotyping a microbial community}",
    journal = {Bioinformatics},
    volume = {37},
    number = {10},
    pages = {1360-1366},
    year = {2021},
    month = {01},
    issn = {1367-4803},
    doi = {10.1093/bioinformatics/btaa977},
    url = {https://doi.org/10.1093/bioinformatics/btaa977},
    eprint = {https://academic.oup.com/bioinformatics/article-pdf/37/10/1360/38663805/btaa977.pdf},
}

Read more on Twitter

License

Hansel and Gretel are distributed under the MIT license, see LICENSE.

Note that the project description data, including the texts, logos, images, and/or trademarks, for each open source project belongs to its rightful owner. If you wish to add or remove any projects, please contact us at [email protected].