Nantes Université

Skip to content

Modifications

Historique de la page
Update Assembly rédigé par Hugo LEFEUVRE's avatar Hugo LEFEUVRE
Afficher les modifications d'espaces
En ligne Côte à côte
Modules/assembly.md
Voir la page @ bd1b7550
--- ---
title: Assembly title: Assembly
--- ---
# Metagenomics assemblies # # Metagenomics assemblies #
Metagenomics assemblies are almost inevitable in metagenomic analysis to reduce the fragmentation of informations and highlight sample characteristics. Reads are assembled into contigs using [megahit](https://github.com/voutcn/megahit) a fast assembler optimized for metagenomes and based on k-mers succinct de Bruijn graphs. Megahit is faster and perform comparably as SPAde and metaSPAdes. SPAde assemblies might be better in term of long contigs and ultra long contigs, it's also show a greater rate of missassemblies which required assembly curation [(1)](https://link.springer.com/article/10.1186/s12864-017-3918-9). Metagenomics assemblies are almost inevitable in metagenomic analysis to reduce the fragmentation of informations and highlight sample characteristics. Reads are assembled into contigs using [megahit](https://github.com/voutcn/megahit) a fast assembler optimized for metagenomes and based on k-mers succinct de Bruijn graphs. Megahit is faster and perform comparably as SPAde and metaSPAdes. SPAde assemblies might be better in term of long contigs and ultra long contigs, it's also show a greater rate of missassemblies which required assembly curation [(1)](https://link.springer.com/article/10.1186/s12864-017-3918-9).
...@@ -20,7 +17,7 @@ For now, misasassemblies are only detected, therefore, if it's required, assembl ...@@ -20,7 +17,7 @@ For now, misasassemblies are only detected, therefore, if it's required, assembl
Single-assembly is a per-sample approach where reads from one sample are assembled into one assembly. This approach is used by the genes_collection module and might be used by the genomes_collection module. Single-assembly is a per-sample approach where reads from one sample are assembled into one assembly. This approach is used by the genes_collection module and might be used by the genomes_collection module.
``` ```
magneto run single-assembly --config target=single_assembly **snakemake.args magneto run assembly --strategies SASB,SACB --profile config/slurm/ **snakemake.args
``` ```
## Co-assembly ## ## Co-assembly ##
...@@ -28,9 +25,11 @@ magneto run single-assembly --config target=single_assembly **snakemake.args ...@@ -28,9 +25,11 @@ magneto run single-assembly --config target=single_assembly **snakemake.args
Co-assembly require the [simka module](Modules/simka). In this case, reads from each samples of a cluster are assembled together. This approach might be used by the genomes_collection module. Co-assembly require the [simka module](Modules/simka). In this case, reads from each samples of a cluster are assembled together. This approach might be used by the genomes_collection module.
``` ```
magneto run assembly --config target=co_assembly **snakemake.args magneto run assembly --strategies CASB,CACB --profile config/slurm/ **snakemake.args
``` ```
Another way to perform co-assembly can be done **without using simka**, but by knowing in advance which samples you want to group together. The information must be specified **at the beginning**, in the samples.yaml file, where all the samples you want to group together will be under **the same sample, each represented by a run.**
----- -----
[Previous - Simka (Module)](Modules/simka) [Previous - Simka (Module)](Modules/simka)
... ...
......