EV Integration Analysis (ERP167546)

This repository contains alignment scripts, metadata, and soft-clip comparison tools associated with the ENA study accession ERP167546, focused on extracellular vesicle (EV) interactions and viral responses in BMDCs and HEK293T cells.

Figure 1. Workflow for EV Integration Analysis in LydenLab Repository.

Caption: This flow chart outlines the processing steps used in the EV integration pipeline. Starting from combined FASTQ files, reads are aligned to reference genomes or viral sequences using Alignment Scripts. BAM files are then subset to genes of interest (Subset Alignments). Soft-clipped bases are extracted from these BAMs (Extract Soft Clipped Bases) and aligned between experimental and control samples (Align Soft Clips B/t Samples). Finally, the percentage of unique soft-clipped reads per comparison is calculated and exported as CSV results (Analyze % of Unique Soft-Clips per Comparison).

🔬 Experimental Overview

This dataset involves multiple experimental conditions across different cell types and treatment protocols. The overall ENA study accession is ERP167546.

BMDCs Experiments [pod5]

Treatment	Submission Date	Run Accessions	Notes
ATEV (EV-treated)	Jan 8th, 2025	ERR14106287, ERR14106288	Two technical replicate flowcells
PBS (control)	Jan 24th, 2025	ERR14208048, ERR14208049, ERR14208050	Three technical replicate flowcells
Virus Ref 1	Feb 16th, 2025	ERR14376133, ERR14376134	Two technical replicate flowcells
Virus Ref 2	Jul 9, 2025	ERR15277392	One technical replicate flowcells

293T Samples [pod5]

Run Accession	Sample Description
ERR15277390	HEK293T treated with PBS
ERR15277391	HEK293T treated with virus_ref_3

BMDCs and 293T Samples [bam]

Sample Description	Run Accession	File Name	Notes
HEK293T treated with virus_ref_3	ERR16121799	PAY43290_combined_Virus_293T.ubam.bam	unaligned bam
HEK293T treated with PBS	ERR16121798	PAY39146_combined_PBS_293T.ubam.bam	unaligned bam
PBS-treated BMDM	ERR16121797	combined_PBS_BMDM.ubam.bam	unaligned bam
EV-treated BMDM	ERR16121796	combined_EV_BMDM.ubam.bam	unaligned bam
Virus-treated BMDM (Virus Ref 1)	ERR16121795	combined_Virus_Treated_1_BMDM.ubam.bam	unaligned bam
Virus-treated BMDM (Virus Ref 2)	ERR16121794	PAY42928_pass_combined_Virus_Treated_2_BMDM.ubam.bam	unaligned bam

---

📁 Repository Structure

ev-integration/
├── FIGURE_S25B/                        # key scripts for figure generation
├── FIGURE_S25C/
├── LydenLab_alignment_scripts/        # SBATCH scripts for alignment to various references
├── LydenLab_pairwise_bam_comparisons/ # Scripts to compare soft-clipped reads across BAMs
├── LydenLab_subset_alignment_scripts/ # Gene Subset-specific alignment jobs
├── metadata/                          # Submission/sample metadata / spreadsheets for all samples
├── references/                        # Contains viral reference sequences (LydenLab_Virus_Ref 1,2,3)

Methods & Versioning

DNA Libraries were prepared with the SQK-LSK114 gDNA Ligation Sequencing Kit and sequenced with FLO-PRO114M flow cells (Oxford Nanopore Technologies). Real-time basecalled reads were produced with MinKNOW Version 24.06.15 and aligned separately with minimap2 v2.28 to the UCSC hg38 reference genome (10.1101/gr.159624.113), UCSC mm10 reference genome and viral vector (LentiGuide-GFP.fa for BMDCs, Addgene# 200961 for the 293T cells). Selected gene coordinates were queried the UCSC genome browser (https://genome.ucsc.edu/cgi-bin/hgGateway). Reads associated with these genes were extracted with samtools v1.21. pysam v0.22.1 was used to extract left and right soft-clipped ends. Soft-clipped ends greater than or equal to 50 bp from either ATEVs or viral vector conditions were aligned to soft-clipped ends from the PBS condition with minimap2 v2.28. The ratio between the number of unaligned soft-clipped ends and total DNA count was then calculated to score gene integration.

Repository Contents & Key Resources

The repository is organized into directories for figure generation, alignment scripts, BAM comparison utilities, subset alignments, metadata, and references. Below is a detailed breakdown.

---

FIGURE_S25B/

Contains data files and scripts used to generate Supplementary Figure S25B.

• LydenLab_EV_comparison_summary_softclip_unaligned_fraction.csv
  CSV summarizing the fraction of unaligned soft-clipped reads for EV vs. PBS comparisons across targeted genes.

---

FIGURE_S25C/

Contains plotting scripts used to generate Supplementary Figure S25C.

• plot_softclip_ecdf.py
  Python script to generate ECDF (Empirical Cumulative Distribution Function) plots of soft-clip read lengths or qualities across experimental conditions.

---

gene_pairwise_bam_comparisons/

Holds outputs from gene-specific pairwise BAM comparisons. Each subfolder corresponds to a gene of interest, containing intermediate and final results.

• Example subfolder: combined_EV_BMDM_vs_combined_PBS_BMDM/H2-Aa/

  • PBS.softclips.fasta — FASTA of soft-clipped PBS reads for the gene.
  • PBS.softclips.mmi — Minimap2 index for PBS softclips.
  • pbs.subsampled.bam / .bai — Subsampled PBS BAM for this gene.
  • virus.subsampled.bam / .bai — Subsampled BAM for EV/virus condition.
  • Virus_vs_PBS.sam — Alignment of virus-treated softclips to PBS softclips.
  • softclip_summary.csv — Summary of match/mismatch and alignment metrics.
  • softclip_combined_EV_BMDM_vs_combined_PBS_BMDM_H2-Aa.out — Raw minimap2 alignment output for the comparison.

• Gene targets include:
  H2-Aa, H2-Ab1, H2-DMb1, H2-Eb2, H2-K1, Tap1
  for mouse MHC genes, with analogous human HLA/TAP targets in other folders.

---

gene_subsets_bam_bai/

Contains sorted and indexed BAM files for targeted alignments to specific genes. Filenames indicate:

• Sample Name (e.g., combined_EV_BMDM)
• Species (human or mouse)
• Gene (e.g., HLA-B, H2-Aa)

Example:

• combined_EV_BMDM.all.sorted_mouse_H2-Aa.bam
• combined_EV_BMDM.all.sorted_mouse_H2-Aa.bam.bai

These files allow focused analysis on loci of interest without processing whole-genome BAMs.

---

LydenLab_alignment_scripts/

Pre-written SLURM sbatch scripts for aligning combined sample FASTQs against multiple references using minimap2.

• Naming convention:
  align_<sample>_<reference>.sbatch
  where <reference> is hg38, mm10, or one of the three viral references (virus_ref_1, virus_ref_2, virus_ref_3).

• Example scripts:

  • align_combined_EV_BMDM_hg38.sbatch — Align EV-treated BMDM to human genome.
  • align_PAY43290_combined_Virus_293T_virus_ref_3.sbatch — Align HEK293T virus sample to virus reference 3.

These scripts produce sorted, indexed BAMs along with alignment statistics.

---

LydenLab_pairwise_bam_comparisons/

Scripts for pairwise BAM analysis focused on soft-clipped read behavior.

• compare_softclips_bam_pair.py — Compares soft-clip retention and re-alignment between virus-treated (or EV-treated) reads and PBS controls.
• summarize_softclip_unaligned_fraction.py — Summarizes the fraction of unaligned reads for each comparison.
• run_softclip_comparisons_sequential_V2.sh — Batch runner for multiple pairwise comparisons.

---

LydenLab_subset_alignment_scripts/

Scripts to align only reads mapping to selected gene subsets.

• subset_alignments_by_gene_V2.sh — Extracts reads for a set of target genes from full BAMs, then aligns them to relevant references.

---

metadata/

Contains ENA submission and sample metadata files for all experiments.

• Run Submission Files:

  • LydenLab_Run_Submission_BMDCs_EV_Treatment.tsv — Accession IDs and submission details for BMDC EV-treated samples.
  • LydenLab_Run_Submission_fastq.tsv — Metadata for combined FASTQ uploads.

• Sample Metadata Files:

  • LydenLab_Sample_Metadata_BMDCs_PBS_Treatment.tsv — PBS control sample details.
  • LydenLab_Sample_Metadata_Virus_Control_2_3.tsv — Virus control (references 2 & 3) sample info.

---

references/

Contains reference FASTA files for viral vectors.

• LydenLab_Virus_Ref/ — Three separate viral reference genomes:
  • virus_ref_1.fa
  • virus_ref_2.fa
  • virus_ref_3.fa

These are used for targeted alignments and soft-clip mapping.

---

Analysis Workflow Overview

1. Align combined FASTQs to the chosen reference genome or viral sequence using scripts in LydenLab_alignment_scripts/.
2. Subset alignments by gene (if needed) using LydenLab_subset_alignment_scripts/.
3. Perform pairwise BAM comparisons with PBS controls using LydenLab_pairwise_bam_comparisons/.
4. Generate summaries and figures using FIGURE_S25B/ and FIGURE_S25C/ scripts.
5. Interpret results using metadata in metadata/ for experimental context.

---

📚 Citation

If you use this repository or data in your work, please cite the originating study linked to ERP167546.

---

For questions, please contact Theo Nelson (thn4005@med.cornell.edu).