sMAG-pipeline.md

Creation of single-sample MAGs

Table of Contents

Description
Requirements
Assembly
Binning
Quality-Estimation
Dereplication

Description

Requirements

• BBmap
• metaSPAdes
• Megahit
• BWA
• sambamba
• metaBAT2
• CheckM
• dRep

Install tools via bioconda:

conda install bbmap spades megahit bwa sambamba metabat2
# built own environment for CheckM, please download database after the installation
conda create -n checkm checkm-genome
# built own environment for dRep
conda create -n drep drep

Assembly

We use metaSPAdes in default mode to create sample-wise assembly:

# for each Sample ($SampleName) with ReadR1 ($Fastq_R1) and ReadR2 ($Fastq_R2) we preform:
spades.py --meta -t 32 \
--pe1-1 ${Fastq_R1} \
--pe1-2 ${Fastq_R2} \
-o Assembly-${SampleName}

Preparation

We use bbmap and bwa to create bam-file for binning:

1. Removal for short contigs bbmap-tools

cd Assembly-${SampleName}
reformat.sh -Xmx20g threads=32 in=contigs.fasta out=contigs-1500bp.fasta fastaminlen=1500

2. Index contings

bwa index contigs-1500bp.fasta

3. Mapping the reads, pipe to sorted bam-file

bwa mem -t 32 contigs-1500bp.fasta ${Fastq_R1} ${Fastq_R2} | \
sambamba view -t 8 -S -f bam /dev/stdin | \
sambamba sort -t 16 -m 120G -o ${SampleName}.bam /dev/stdin

Binning

runMetaBat.sh -t 32  contigs-1500bp.fasta ${SampleName}.bam

CheckM

# activate CheckM environment
conda activate checkm
checkm lineage_wf -f checkM-SCG-single.txt \
-t 32 --pplacer_threads 10 -x fa . \
./checkM-out-single

Dereplication

Create first a checkM-data.txt including all CheckM information of all MAGs you want to cluster:

genome,completeness,contamination,strain_heterogeneity
iMGMC-100.fa,97.07,1.27,1.27
iMGMC-1000.fa,58.9,0.67,0.67
iMGMC-1001.fa,79.84,0.23,0.23
[...]

Run dRep with different settings using all mMAGs (comp >= 50, con < 10) located in the folder "all5010"

# activate dRep environment
# run for mMAGs
conda activate dRep
dRep dereplicate drep-outout-SkipSecondary-ANI95-c50c10 \
--genomeInfo checkM-data.txt \
-comp 50 -con 10 \
--P_ani 0.95 --SkipSecondary -p 32 \
-g all5010/*.fa


# run for hqMAGs
conda activate dRep
dRep dereplicate drep-outout-SkipSecondary-ANI95-c50c10 \
--genomeInfo checkM-data.txt \
-comp 90 -con 5 \
--P_ani 0.95 --SkipSecondary -p 32 \
-g all5010/*.fa


# default run without secondary clustering
# Minumum genome completeness (default: 75)
conda activate dRep
dRep dereplicate drep-outout-SkipSecondary-ANI95-c50c10 \
--genomeInfo checkM-data.txt \
--P_ani 0.95 --SkipSecondary -p 32 \
-g all5010/*.fa

Pipeline to create integrated MAGs:

Code for assembly, binning and 16S rRNA gene reconstruction

Linking of RAMBL sequences to bins:

Code for linking 16S rRNA genes to bins