SMART-PTA

Nextflow pipeline

in support of the preprint: Large scale single-cell phylogenetic mapping of clonal evolution in the human aging esophagus

Quick start & Test run

   git clone https://github.com/landau-lab/smartpta.git && cd smartpta

   nextflow run workflows/scVC.nf -profile test -stub

   #explore example output
   tree -C output

Installation

Requirements

These workflows were developed using the following:

• Nextflow 22.10.4+
• Singularity 3.8.6+

The pipelines will automatically pull the required containers when run. Our reference data bundle is required in order to run the pipelines, in order to download the required files run:

   cd resources
   ./ref_setup.sh

More information on the reference data bundle can be found here.

Usage

scVariantCalling (scVC.nf)

This will run the following steps:

• Duplicate Marking
• Contamination Estimate
• UG DeepVariant
• GLNexus joint genotyping
• Variant Annotation

Create a bam list

   #e.g. bam_list.txt
   /path/to/bam1.bam
   /path/to/bam2.bam
   /path/to/bam3.bam

   nextflow workflows/scVC.nf --bam_list <bam_list> --sample_id <sample_id>

A version of the pipeline targeting Illumina data is also availible workflows/scVC-il.nf. The input of this version is a list of paired fastq files.

   #e.g. fq_list.txt
   /path/to/fastq1_R1.fastq.gz /path/to/fastq1_R2.fastq.gz
   /path/to/fastq2_R1.fastq.gz /path/to/fastq2_R2.fastq.gz
   /path/to/fastq3_R1.fastq.gz /path/to/fastq3_R2.fastq.gz

   nextflow workflows/scVC-il.nf --fq_list <fq_list> --sample_id <sample_id>

scRNAseq Analysis (scRNA.nf)

This will run the following steps:

• Quality control with FastP
• Alignment with STAR
• Quantification with HTSeq
• Merging of counts
• Quality control report with MultiQC

Create an RNA-seq fastq list

   #e.g. rna_fastq_pairs.txt
   /path/to/fastq1_R1.fastq.gz /path/to/fastq1_R2.fastq.gz
   /path/to/fastq2_R1.fastq.gz /path/to/fastq2_R2.fastq.gz
   /path/to/fastq3_R1.fastq.gz /path/to/fastq3_R2.fastq.gz

nextflow workflows/scRNA.nf --rna_fastq_table <rna_fastq_pairs> --sample_id <sample_id>

Config overrides

It might be necessary to override the default config for certain processes. For example, on our cluster, our older cpus do not support AVX512 instructions, this makes running UGDeepVariantCPU very slow when allocated to these nodes, so we need to override the default cluster options to select nodes with newer CPU architectures via thier feature flags (herev5|v6 selects nodes with AVX512 support).

process {
    ...
    withName: UGDeepVariantCPU { clusterOptions = '-C "v5|v6"' }
}

Citation

@article{Prieto2025,
  author = {Prieto, Tamara and Yuan, Dennis J and Zinno, John and Hughes, Clayton and Midler, Nicholas and Kao, Sheng and Huuhtanen, Jani and Raviram, Ramya and Fotopoulou, Fenia and Ruthen, Neil and Rajagopalan, Srinivas and Schiffman, Joshua S and D Avino, Andrew R and Yoon, Sang-Ho and Sotelo, Jesus and Omans, Nathaniel D and Wheeler, Noelle and Garces, Alejandro and Pradhan, Barun and Cheng, Alexandre Pellan and Robine, Nicolas and Potenski, Catherine and Godfrey, Katharine and Kakiuchi, Nobuyuki and Yokoyama, Akira and Ogawa, Seishi and Abrams, Julian and Raimondi, Ivan and Landau, Dan A},
  title = {Large-scale single-cell phylogenetic mapping of clonal evolution in the human aging esophagus},
  year = {2025},
  doi = {10.1101/2025.10.11.681805},
  journal = {bioRxiv}
}