Bio Read Qc Contamination Screening

# Database locations
DATABASE	Human	/path/to/human/genome
DATABASE	Mouse	/path/to/mouse/genome
DATABASE	Ecoli	/path/to/ecoli/genome
DATABASE	PhiX	/path/to/phix/genome
DATABASE	Adapters	/path/to/adapters
DATABASE	rRNA	/path/to/rrna

# Aligner (bowtie2 recommended)
BOWTIE2	/path/to/bowtie2

# Or use BWA
# BWA	/path/to/bwa

# Threads
THREADS	8

Pre-built Databases

# Download common screening databases
fastq_screen --get_genomes

# Downloads to ~/fastq_screen_databases/
# Includes: Human, Mouse, Rat, E.coli, PhiX, Adapters, etc.

Screening Options

# Number of reads to sample (default 100000)
fastq_screen --subset 200000 sample.fastq.gz

# Use all reads (slow)
fastq_screen --subset 0 sample.fastq.gz

# Set threads
fastq_screen --threads 8 sample.fastq.gz

# Paired-end (screen R1 only by default)
fastq_screen sample_R1.fastq.gz

# Force screening both pairs
fastq_screen --paired sample_R1.fastq.gz sample_R2.fastq.gz

Output Options

# Generate PNG plot (default)
fastq_screen sample.fastq.gz

# No plot (text only)
fastq_screen --nograph sample.fastq.gz

# Generate additional mapping statistics
fastq_screen --tag sample.fastq.gz

# Filter reads by mapping (keep unmapped to all genomes)
fastq_screen --filter 0000 sample.fastq.gz

# Keep only reads mapping to first genome (e.g., Human)
fastq_screen --filter 1--- sample.fastq.gz

Filter Codes

Use --filter to select reads based on mapping status:

# Example: Keep reads mapping only to Human (first genome)
# Human:1, all others:0
fastq_screen --filter 10000 sample.fastq.gz

# Keep reads NOT mapping to anything (clean reads)
fastq_screen --filter 00000 sample.fastq.gz

Output Files

Results Format

#Fastq_screen version: 0.15.3
Genome	#Reads_processed	#Unmapped	%Unmapped	#One_hit_one_genome	%One_hit_one_genome	#Multiple_hits_one_genome	%Multiple_hits_one_genome	#One_hit_multiple_genomes	%One_hit_multiple_genomes	Multiple_hits_multiple_genomes	%Multiple_hits_multiple_genomes
Human	100000	2000	2.00	95000	95.00	1000	1.00	1500	1.50	500	0.50
Mouse	100000	98000	98.00	100	0.10	50	0.05	1500	1.50	350	0.35

Interpreting Results

Expected Results by Sample Type

Common Issues

MultiQC Integration

FastQ Screen results are automatically detected by MultiQC:

# Screen all samples
for f in *.fastq.gz; do
    fastq_screen --outdir screen_results/ "$f"
done

# Aggregate with MultiQC
multiqc screen_results/

Custom Database Setup

Create Bowtie2 Index

# Index a FASTA file
bowtie2-build reference.fa reference

# Add to config
# DATABASE	MyGenome	/path/to/reference

Common Databases to Include

Example Workflows

Standard Screening

# Download databases
fastq_screen --get_genomes

# Screen samples
fastq_screen --outdir screen_results/ --threads 8 *.fastq.gz

# Check results
multiqc screen_results/

Remove Contamination

# Screen and tag reads
fastq_screen --tag sample.fastq.gz

# Filter to keep only Human reads (assuming Human is first database)
fastq_screen --filter 3----- --tag sample.fastq.gz

# Or use BBDuk for removal
bbduk.sh in=sample.fastq.gz out=clean.fastq.gz \
    ref=contaminants.fa k=31 hdist=1

Related Skills

• quality-reports - FastQC shows overrepresented sequences
• adapter-trimming - Remove adapter contamination
• metagenomics - Deeper taxonomic analysis