Skip to content

Latest commit

 

History

History
192 lines (136 loc) · 5.57 KB

README.md

File metadata and controls

192 lines (136 loc) · 5.57 KB
Raw

Linking of reconstructed 16S rRNA genes to bins

Table of Contents

Description
Requirements
Blast-search
Split-Mapping
Correlation
Integration
Evaluation

Description

linking

Requirements

Blast-search

1. Create Blast database with reconstructed 16S rRNA sequences from RAMBL

makeblastdb -type nucl -in RAMBL-16SrRNAsequences.fasta

2. Run Blast for all bins ($BinFastaFile)

blastn \
-query $BinFastaFile \
-db RAMBL-16SrRNAsequences.fasta \
-max_target_seqs 50 \
-outfmt 6 \
-evalue 0.000001 \
-num_threads 20 \
-out blastn-bins_Finals-full-nr99.50-hits-fasta.outfmt6

3. Filter Blast results (min ident 97% and min coverage 100bp)

cat blastn-bins_Finals-full-nr99.50-hits-fasta.outfmt6 | \
awk '$3 > 97' | awk '$4 > 100' | cut -f1,2,12 > filtered-BlastOut.tab

4. Create Blast results matrix

Needed Input files:

  • filtered-BlastOut.tab
  • list for all 16SrRNA names in text-format: "16SrRNAnames.txt"
  • list for all bin names in text-format: "BinNames.txt"

Run reformat script (iMGMC/linking/reformat-blast-results.sh)

reformat-blast-results.sh filtered-BlastOut.tab 16SrRNAnames.txt BinNames.txt

#final output: "matrix-Blast.tab"

Split-Mapping

1. Create BBmap index for all bins

bbsplit.sh ref=${Folder-of-all-bins}

2. Run BBsplit to split each library into reads mapping to each bin

# for each Sample ($SampleName) with ReadR1 ($Fastq_R1) and ReadR2 ($Fastq_R2) we preform:
mkdir splited-read-${SampleName}
bbsplit.sh -Xmx80g usejni=t unpigz=t threads=24 minid=0.90 \
ambiguous2=toss \
in=${SampleName}_R1_rmhost.fastq.gz \
in2=${SampleName}_R2_rmhost.fastq.gz \
basename=./splited-read-${SampleName}/out_%.fastq.gz

3. Map all mapped reads of a bin to all 16S rRNA sequences from RAMBL

mkdir statsfiles-unpaired-all scafstats-statsfiles-unpaired-all \
cov-statsfiles-unpaired-all rpkm-statsfiles-unpaired-all
# for each bin ($BinNr) we preform:
zcat ./splited-read-*/out_bin_${BinNr}.fastq.gz | \
bbmap.sh local=t -Xmx30g unpigz=t threads=${usedCores} minid=0.90 \
ref=RAMBL-16sRNAsequences.fasta nodisk \
interleaved=false \
statsfile=./statsfiles-unpaired-all/${BinNr}.statsfile \
scafstats=./scafstats-statsfiles-unpaired-all/${BinNr}.scafstats \
covstats=./cov-statsfiles-unpaired-all/${BinNr}.covstat \
rpkm=./rpkm-statsfiles-unpaired-all/${BinNr}.rpkm \
sortscafs=f nzo=f ambiguous=all \
in=stdin.fastq

3. Summarize statistics

# run the script in the same directory
# it use the dir "statsfiles-unpaired-all"
make-splitmapping-stats.sh

#final output: "final-unambiguousReads.tab"

Correlation

1. Create 16S rRNA gene abundances over all samples

# for each Sample ($SampleName) with ReadR1 ($Fastq_R1) and ReadR2 ($Fastq_R2) we preform:
bbmap.sh -Xmx30g unpigz=t threads=${usedCores} minid=0.90 \
ref=RAMBL-16sRNAsequences.fasta nodisk \
statsfile=16SrRNA-statsfiles/${SampleName}.statsfile \
scafstats=16SrRNA-scafstats-statsfiles/${SampleName}.scafstats \
covstats=16SrRNA-cov-statsfiles/${SampleName}.covstat \
rpkm=16SrRNA-rpkm-statsfiles/${SampleName}.rpkm \
sortscafs=f nzo=f ambiguous=all local=t \
in=${SampleName}_R1_rmhost.fastq.gz \
in2=${SampleName}_R2_rmhost.fastq.gz

2. Summarize statistics and create matrix file for 16S rRNA abundances

# run the script in the same directory
# it use the dir "scafstats-statsfiles"
make-16S-mapping-stats.sh

#final output: "16S-abundances.tab"

3. Create bin abundances over all samples

# create conducted bin-fasta file from all bins in ${Folder-of-all-bins}
mkdir bin-abundances
cd bin-abundances
for binfile in ${Folder-of-all-bins}
do
BinFileName=${binfile##*/}
echo ">${BinFileName%.*}" >> bin-sequences.fasta
fgrep -v ">" $binfile >> bin-sequences.fasta
done

# create ref for bbmap
bbmap.sh ref=bin-sequences.fasta

# for each Sample ($SampleName) with ReadR1 ($Fastq_R1) and ReadR2 ($Fastq_R2) we preform:
bbmap.sh -Xmx30g unpigz=t threads=${usedCores} minid=0.90 \
statsfile=bin-statsfiles/${SampleName}.statsfile \
scafstats=bin-scafstats-statsfiles/${SampleName}.scafstats \
covstats=bin-cov-statsfiles/${SampleName}.covstat \
rpkm=bin-rpkm-statsfiles/${SampleName}.rpkm \
sortscafs=f nzo=f ambiguous=all \
in=${SampleName}_R1_rmhost.fastq.gz \
in2=${SampleName}_R2_rmhost.fastq.gz

4. Summarize statistics and create matrix file for bin abundances

# run the script in the same directory
# it use the dir "ref-statsfiles"
make-bin-mapping-stats.sh

#final output: "bin-abundances.tab"

Integration

Needed files:

  • matrix-Blast.tab
  • bin-abundances.tab
  • 16S-abundances.tab
  • final-unambiguousReads.tab
  • BinMetaData.csv ("BinID""genome size bp")

Create bin metadata file

echo -e "BinID\tgenome size bp" > BinMetaData.csv
for bin-file in ${Folder-of-all-bins}
do
BinFileName=${bin-file##*/}
BinName=${BinFileName%%.*}
fgrep -v ">" $bin-file | wc -m | sed -e "s/^/${BinName}\t/" >> BinMetaData.csv
done

Run Integration pipeline in R

Rscript workflow_management_16S_meta_int.R \
bin-abundances.tab \
BinMetaData.csv \
16S-abundances.tab \
final-unambiguousReads.tab \
blast-matrix.txt \
${PWD}

Evaluation

Evaluation with simulated data