🎯 Ultrafast Short-Read Genomic Alignment
Bowtie 2 combines the memory efficiency of the Ferragina-Manzini (FM) index with hardware-accelerated dynamic programming. It excels at aligning 50-1000 bp reads for gDNA, ChIP-seq, ATAC-seq, and metagenomic host decontamination.
🎯 Alignment Mode & Minimum Score Calculator
Interactive ToolPreset & Read Specifications
Calculated Parameters
🦠 Metagenomic Host Depletion Simulator
Interactive FilterIn microbiome sequencing (saliva, biopsy, gut), 70-95% of reads come from the human host. Bowtie2 can output unmapped read pairs directly into clean FASTQ files with zero loss:
# Filter out host (human) reads; keep only pure microbial reads bowtie2 -p 8 -x hg38_index/genome \ -1 raw_R1.fastq.gz -2 raw_R2.fastq.gz \ --un-conc-gz microbial_reads_R%.fastq.gz \ --very-sensitive \ -S /dev/null
0. Burrows-Wheeler Transform & FM-Index Concepts
Bowtie 2 creates an FM-index from the Burrows-Wheeler Transformed reference genome. This allows matching 22-nucleotide seeds in exact algorithmic time proportional only to the read length, completely independent of the reference genome size.
1. Indexing Reference Genome (.bt2 vs .bt2l)
Genomes < 4 GB use standard 32-bit .bt2 indexes; larger genomes automatically use 64-bit .bt2l indexes.
# Build Bowtie 2 index using 8 threads bowtie2-build --threads 8 reference.fasta bt2_index/genome
2. gDNA & Exome Alignment Execution
Standard paired-end gDNA alignment streaming directly to coordinate-sorted BAM.
bowtie2 -p 8 --sensitive \ -x bt2_index/genome \ -1 sample_R1.fastq.gz -2 sample_R2.fastq.gz | \ samtools sort -@ 8 -m 2G -o sample_aligned.bam - samtools index -@ 8 sample_aligned.bam
3. ChIP-Seq & ATAC-Seq Short Read Mapping
ChIP-seq and ATAC-seq require strict concordant mapping and fragment size filters (e.g. -X 2000 for wide nucleosome fragments).
# ATAC-seq mapping allowing fragment inserts up to 2000 bp bowtie2 -p 8 --very-sensitive -X 2000 --no-mixed --no-discordant \ -x bt2_index/genome \ -1 atac_R1.fastq.gz -2 atac_R2.fastq.gz | \ samtools view -@ 8 -b -q 30 -F 4 -F 256 - | \ samtools sort -@ 8 -o atac_clean.bam -
4. End-to-End vs Local Soft-Clipping
In --local mode, Bowtie 2 allows the ends of reads to be soft-clipped if they contain residual adapter sequences or poor quality basecalls.
# Local mode for trimming-free robust alignment bowtie2 -p 8 --local --very-sensitive-local \ -x bt2_index/genome \ -1 sample_R1.fastq.gz -2 sample_R2.fastq.gz -S local_aligned.sam
5. Multi-mapping & Discordant Read Handling
Control whether Bowtie 2 reports one primary alignment, searches for up to $k$ valid alignments (-k 10), or searches for all valid alignments (-a).
# Report up to 5 distinct alignments per read for repetitive region inspection bowtie2 -p 8 -k 5 \ -x bt2_index/genome \ -U single_reads.fastq.gz -S multi_hits.sam
6. Metagenomic Host Read Depletion
Decontaminate human/mouse host reads from environmental metagenomes prior to de novo assembly.
bowtie2 -p 8 --very-sensitive \ -x host_genome/hg38 \ -1 sample_R1.fq.gz -2 sample_R2.fq.gz \ --un-conc-gz decontaminated_R%.fq.gz \ --al-conc-gz host_contaminant_R%.fq.gz \ -S /dev/null
7. Alignment Score Interpretation (AS:i vs XS:i)
Bowtie 2 records the best alignment score in AS:i and the second-best in XS:i. If AS == XS, the read is a multi-mapper (MAPQ = 0 or 1).
# Filter out multi-mapping reads: keep only unique alignments (MAPQ >= 20) samtools view -@ 4 -b -q 20 sample_aligned.bam -o uniquely_mapped.bam
8. Downstream ChIP-Seq Peak Calling (MACS3)
Call narrow or broad binding summits using MACS3 with ChIP treatment and Input control BAMs.
# Call narrow histone / transcription factor peaks macs3 callpeak \ -t chip_treatment.bam \ -c chip_control_input.bam \ -f BAMPE -g hs -n sample_peaks \ -q 0.05 --outdir macs3_peaks/
🔀 Decision Architecture: Short Read Aligner Choice
Decision TreeBowtie 2 is the gold standard for ChIP-seq and ATAC-seq because of its fine control over discordant pairs (
--no-discordant) and fragment insert size limits (-X).