libseq

Nextflow pipeline for processing paired-end Illumina reads, especially useful for analysing amplicons of library variants. Takes raw FASTQ files and outputs per-bin sequence frequency counts.

Pipeline steps

1. fastp — quality control and adapter trimming
2. FLASH2 — merge paired-end reads by overlap
3. cutadapt — orient reads using inline barcodes and a 3′ anchor sequence
4. cutadapt — demultiplex into per-bin FASTQ files using inline barcodes
5. cutadapt (optional) — extract the library region between left and right flanking sequences; reads that are untrimmed, too short, or too long are written to separate files for inspection
6. Python — count the frequency of each unique sequence per bin
7. MultiQC — aggregate QC report across all steps

Outputs

• 06_counts/ — per-bin sequence counts for the extracted library region (TSV)
• 07_merged_counts/ — all bins merged into a single TSV
• 06_whole_counts/ — per-bin sequence counts for full demuxed reads (TSV)
• 07_merged_whole_counts/ — all bins merged into a single TSV
• 00_reports/ — MultiQC HTML report

Whole-read and library-region counting can each be toggled independently (see parameters below).

Requirements

• Java
• Nextflow
• conda

Usage

Clone and run locally:

git clone https://github.com/ilyacarey/libseq.git
cd libseq
nextflow run main.nf -profile conda

Run directly from GitHub (barcodes FASTA must be available locally):

nextflow run ilyacarey/libseq -profile conda

Update the relevant parameters in nextflow.config before running, or pass them on the command line:

nextflow run main.nf -profile conda \
  --read1 sample_R1.fastq.gz \
  --read2 sample_R2.fastq.gz \
  --sample my_sample \
  --barcodes_fasta barcodes_anchored.fasta

Key parameters

Parameter	Description
read1 / read2	Input paired-end FASTQ files
sample	Sample name used for output file naming
barcodes_fasta	FASTA of inline barcodes (must be 5′-anchored with ^)
orient_3p_anchor	3′ anchor sequence for read orientation (anchor to end with $)
left_flank	Sequence immediately upstream of the library region
right_flank	Sequence immediately downstream of the library region
lib_min_len / lib_max_len	Expected length range (bp) of the extracted library region
count_library	Output counts of extracted library region (default: true)
count_whole_reads	Output counts of full demuxed reads (default: true)
cleanup	Delete work/ directory on successful completion (default: true); set to false to enable -resume
cpus	Number of CPU threads (default: 8)