ht‑selex‑pipeline

Snakemake workflow for end‑to‑end processing of high‑throughput SELEX (HT‑SELEX) sequencing data.
It performs adapter trimming, sequence counting, enrichment analysis versus Round 0, motif discovery with MEME, and automated visualisation, producing publication‑ready outputs from raw FASTQ files in a single command.

---

Background & Rationale

High‑throughput SELEX iteratively enriches nucleic‑acid libraries for high‑affinity binders (aptamers). Each round is deep‑sequenced, creating gigabytes of reads that must be:

1. Quality‑checked and adapter‑trimmed to avoid bias.
2. Collapsed to unique sequences and counted per round.
3. Quantitatively compared to the naïve pool (Round 0) to identify enriched candidates.
4. Interrogated for conserved motifs that underpin binding.
5. Visualised for rapid assessment and downstream decision‑making.

Manual, ad‑hoc analysis is error-prone and non-reproducible. This pipeline encapsulates best‑practice steps in an automated, version‑controlled workflow that scales from laptops to HPC clusters.

---

Workflow Overview

graph TD
  A("Raw FASTQ (per round)") --> B("FastQC")
  B --> C("Cutadapt – trim adapters")
  C --> D("Count unique sequences")
  D --> E("Enrichment vs Round 0")
  E --> F("MEME motif discovery")
  E --> G("Enrichment plot (PNG)")

Loading

Step	Tool / Script	Key Output
QC	FastQC	qc/*.zip (HTML inside)
Trim	Cutadapt	trimmed/*.trimmed.fastq.gz
Count	scripts/count_seqs.py	counts/*.tsv
Enrich	scripts/calc_enrichment.py	enrichment/enrichment_R*.tsv
Motif	MEME‑Suite (meme)	motifs/R*/* (HTML + logos)
Plot	scripts/plot_enrichment.py	plots/enrichment_R*.png

All dependencies are installed into an isolated micromamba environment via environment.yml.

---

Quick Start

# 1. Clone the repo
git clone https://github.com/Lambda-Epsilon/ht-selex-demo.git
cd ht-selex-pipeline

# 2. Create & activate environment
micromamba env create -f environment.yml
micromamba activate htselex

# 3. Edit config.yaml
#    - datasets: experiment IDs
#    - rounds:   sequencing rounds available
#    - runs:     SRR accessions for automatic download (optional)

# 4. (Optional) Download raw FASTQs
./download_data.sh        # uses runs: from config.yaml

# 5. Run the pipeline on 32 logical cores
snakemake --cores 32 --rerun-incomplete --latency-wait 60

Outputs appear under results/<dataset>/.

---

Results (Example demonstration, dataset DRA009383):

Dataset	Rounds analysed	Top enriched sequence	Enrichment (R*)	Leading motif (E‑value)
DRA009383	0–1	GCCAGTAG…	7532.0×	AACTTCGAAAG (8.2e^-162)

Full QC metrics, enrichment tables, motif HTML reports and PNG plots are provided in the results/ directory.

---

File Tree (after successful run)

results/
└── DRA009383
    ├── qc/
    ├── trimmed/
    ├── counts/
    ├── enrichment/
    ├── motifs/
    │   └── R1/
    └── plots/
        └── enrichment_R1.png

---

Customisation

Task	How to modify
Change adapter sequence	Edit adapter: in config.yaml
Add extra rounds	Append to rounds: array
Run on cluster	Provide a Snakemake profile or --cluster directive
Increase per‑rule threads	Adjust THREADS in config.yaml and rule threads: fields
Extend motif parameters	Edit the meme rule in Snakefile