Skip to content

Latest commit

 

History

History
74 lines (56 loc) · 2.49 KB

File metadata and controls

74 lines (56 loc) · 2.49 KB

Contamination Screening - Usage Guide

Overview

FastQ Screen quickly identifies contamination by aligning a subset of reads against multiple reference genomes. It detects cross-species contamination, bacterial contamination, adapter sequences, and sample swaps.

Prerequisites

conda install -c bioconda fastq-screen

# Download pre-built databases
fastq_screen --get_genomes

Quick Start

Tell your AI agent what you want to do:

  • "Screen my FASTQ files for contamination"
  • "Check if my human sample has any mouse contamination"
  • "Verify my samples are not swapped"

Example Prompts

Basic Screening

"Run FastQ Screen on my samples to check for contamination"

"Screen my FASTQ files against human, mouse, and common contaminants"

Troubleshooting

"My RNA-seq has unexpected results, check for contamination or sample swap"

"Identify the source of contamination in my samples"

Configuration

"Set up FastQ Screen with custom genome databases"

"Add E. coli and yeast to my contamination screening panel"

What the Agent Will Do

  1. Configure FastQ Screen with appropriate reference databases
  2. Run screening on a subset of reads from each sample
  3. Generate reports showing percentage mapping to each genome
  4. Identify problematic samples with contamination or swaps
  5. Recommend appropriate remediation steps

Interpretation Guide

Good Results (Human Sample)

Genome      %One_hit  %Multi_hit
Human       95.0      2.0
Mouse       0.1       0.0
Ecoli       0.0       0.0
Adapters    0.1       0.0

Common Issues

Issue Pattern Solution
Bacterial contamination High E.coli/bacteria % Investigate source
Sample swap Wrong species dominant Re-sequence or exclude
Adapter contamination High Adapters % Run adapter trimming
PhiX spike-in High PhiX % Filter PhiX reads
rRNA contamination High rRNA % rRNA depletion failed

Tips

  • Run contamination screening early before investing in alignment and analysis
  • Include common contaminants: E. coli, yeast, PhiX, adapters
  • For xenograft samples, expect mixed human/mouse signal
  • Sample swaps are easier to detect than fix - verify sample identity early
  • Use --subset to screen a random subset of reads for faster results

Resources