Evaluation of a high-throughput, cost-effective Illumina library preparation kit

Eric S. Tvedte

2020-07-07

The repository contains Supplementary Data for the manuscript, including Tables, Figures, and Files.

Table of Contents

1. Read representation
2. Riptide contamination analysis
3. Read mapping performance and genome coverage analysis
4. De novo assembly
5. Metagenome analysis
6. System requirements
7. Installation
8. GitHub repository contents

Read representation

Raw read counts
Run scripts/Riptide.FileS1.Rmd using sample_data_files/raw_reads.txt

Riptide contamination analysis

Filter 1: Species assignment, retain primary reads Map reads to reference containing combined genomes from all organisms in Riptide experiment
bwa mem -k 23 combined.reference.fasta R1.fastq.gz R2.fastq.gz | samtools view -bho mapped.bam -
Sort BAM with samtools
samtools sort -m 2G -o sorted.bam mapped.bam
Mark duplicates with PicardTools
java -Xmx10g -jar picard.jar MarkDuplicates I=sorted.bam O=dupsmarked.bam M=dupsmetrics.txt
Assign reads to species
samtools view -F 4 -F 256 -F 1024 -F 2048 -bho primary.bam dupsmarked.bam
samtools view primary.bam | awk -F "\t" '{print $3}' | sort -n | sed "s/_scaf.*//g" | uniq -c | awk '{print $2"\t"$1}' > speciesmap.txt
Species assignment files were compiled for all libraries, with an additional row added containing numbers corresponding to the well position of each sample on the Riptide plate
Visualization
Run scripts/Riptide.FileS4 using sample_data_files/contamination.txt

Filter 2: Species assignment, retain primary reads with MAPQ > 20 samtools view -h -F 4 -F 256 -F 1024 -F 2048 -q 20 dupsmarked.bam | samtools view -bh - > filter_lowMAPQ.bam
samtools view filter_lowMAPQ.bam | awk -F "\t" '{print $3}' | sort -n | sed "s/_scaf.*//g" | uniq -c | awk '{print $2"\t"$1}' > speciesmap.txt
Compilation of data and visualization was performed as described above

Genome coverage analysis

Mapped read counts
Subsample read sets
seqkit sample -s 13 -j 4 -p #### -o R1.subsample.fastq R1.fastq.gz -p parameter is dependent on readset

Map reads with bwa
bwa mem -k 23 reference.fasta R1.subsample.fastq R2.subsample.fastq | samtools view -bho mapped.bam
Sort BAM with samtools
samtools sort -m 2G -o sorted.bam mapped.bam
Mark duplicates with PicardTools
java -Xmx10g -jar picard.jar MarkDuplicates I=sorted.bam O=dedup.bam M=dupsmetrics.txt VALIDATION_STRINGENCY=SILENT AS=true CREATE_INDEX=true REMOVE_DUPLICATES=true samtools view -F 4 -F 256 -F 1024 -F 2048 -bho primary.bam dedup.bam Visualize counts
Run scripts/Riptide.FileS1.Rmd using sample_data_files/mapped_reads.txt

Histograms of genome breadth and depth of coverage
Make genomebed file
awk -v OFS='\t' {'print $1, $2'} genome.fasta.fai > genomefile.bed
Make non-overlapping sliding window intervals using bedtools
bedtools makewindows -g genomefile.bed -w 1000 > genome_windows.bed
Determine genome-wide coverage using bedtools
bedtools coverage -a genome_windows.bed -b primary.bam -hist | grep ^all > genome_coverage.bed
*Visualize genome-wide coverage Run scripts/Riptide.FileS2.Rmd - E.coli using sample_data_files/Ecoli_Riptide_depth.bed and Ecoli_KAPA_depth.bed

GC bias assessment
java -jar picard.jar CollectGcBiasMetrics I=primary.bam O=gcbias_metrics.txt CHART=gcbias_metrics.pdf S=summary_metrics.txt R=reference_sequence.fasta
grep -v '#' gcbias_metrics.txt | awk '{print $3"\t"$4"\t"$7}' > GC.bias.final.out

Visualize depth vs. GC Run scripts/Riptide.FileS3.Rmd using sample_data_files/Acinetobacter_depthvsGC.txt

De novo assemblies

De novo assembly of readsets with SPAdes
spades.py -o SPAdes_output_dir --pe1-1 R1.subsample.fastq --pe1-2 R2.subsample.fastq -m 200

Assembly evaluation
Contiguity metrics with QUAST
quast.py -o $f/quast_assembly_stats -m 1000 contigs.fasta
Prokaryotes: conserved bacteria genes with BUSCO
python run_busco.py -f -c 8 -t /local/scratch/etvedte/tmp -i contigs.fasta -o busco_output_dir -l bacteria_odb9 -m geno
Eukaryotes: conserved metazoa genes with BUSCO
python run_busco.py -f -c 8 -t /local/scratch/etvedte/tmp -i contigs.fasta -o busco_output_dir -l metazoa_odb9 -m geno

Metagenome analysis

Raw read counts for metagenome libraries Visualize counts
Run scripts/Riptide.FileS6.Rmd using sample_data_files/metagenome_data.txt

Assignment of LCAs to reads with kraken2
Build kraken2 library
kraken2-build --download-taxonomy --db kraken2-nt
kraken2-build --download library nt --db kraken2-nt
kraken2-build --build --db kraken2-nt Assign reads using kraken2
kraken2 --threads 16 --db kraken2-nt --paired --report kraken2_report.txt R1.fastq.gz R2.fastq.gz Retrieval of domain assignments for reads
awk '$4=="D"' kraken2_report.txt
Retrieval of phylum assignments for reads, minimum percentage cutoff of 0.5
awk '$4=="P"' kraken2_report.txt | awk '$1>0.5' -
Visualization Run scripts/Riptide.FileS6.Rmd using sample_data_files/metagenome_data.txt

System requirements

R scripts were run using Windows 10 x64 with RStudio v1.1.463 using this R session:

R version 3.5.1 (2018-07-02)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows 10 x64 (build 18362)

Installation

Requirements for installed packages are available in Supplementary Files.

GitHub repository contents

supplementary_figures
Contains Supplementary Figures reported in the manuscript, Figure S1 - S14

supplementary_tables
Contains Supplementary Tables reported in the manuscript, Table S1 - S13

scripts
Contains six Rmd scripts used for data analysis:

Riptide.FileS1: Read composition, mapping performance, and indel quantification

Riptide.FileS2: Read coverage histograms

Riptide.FileS3: Contamination analysis

Riptide.FileS4: Visualizing breakpoints in E. coli genome assemblies

Riptide.FileS5: Metagenome analysis

Sample input files are provided in sample_data_files. Input and output file paths are hard-coded in the scripts, change these to run the scripts on your local system.

htmls
This folder contains output html files generated from Rmd files using the R package knitr.

sample_data_files
This folder contains sample input files for data analysis.

raw+mapped_reads.txt: for use with Riptide.FileS1.Rmd

Ecoli_Riptide_depth.bed, Ecoli_KAPA_depth.bed, Shigella_KAPA_depth.bed, Shigella_KAPA_depth.bed: for use with Riptide.FileS2.Rmd

Riptide_contamination_filter1.txt: for use with Riptide.FileS3.Rmd

12 files with .coords extension: for use with Riptide.FileS4.Rmd

metagenome_data.txt: for use with Riptide.FileS5.Rmd