PathogenTrack

PathogenTrack is an unsupervised computational software that uses unmapped single-cell RNAseq reads to characterize intracellular pathogens at the single-cell level. It is a python-based script that can be used to identify and quantify intracellular pathogenic viruses and bacteria reads at the single-cell level. PathogenTrack has been tested on various scRNA-seq datasets derived from simulated and real datasets and performed robustly. The detailes are described in our paper Decoding Intracellular Pathogens of scRNA-seq experiments with PathogenTrack and SCKIT.

System Requirements

PathogenTrack has been tested on Linux platform with CentOS 7 operation system. The RAM is 120 GB, with 40 computational threads.

Installation

PathogenTrack can be installed in two steps:

1 . Installing Miniconda on Linux Platform. For details, please refer to Miniconda Installation.

wget https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh
bash Miniconda3-latest-Linux-x86_64.sh

2 . Installing PathogenTrack.

conda env create -f environment.yml

Users can install the dependencies manually. The dependencies and test versions are listed below.

Databases Preparation

1. Prepare the Human genome database

Download the Human GRCh38 genome and genome annotation file, and then decompress them:

wget ftp://ftp.ensembl.org/pub/release-101/fasta/homo_sapiens/dna/Homo_sapiens.GRCh38.dna.toplevel.fa.gz
gzip -d Homo_sapiens.GRCh38.dna.toplevel.fa.gz
wget ftp://ftp.ensembl.org/pub/release-101/gtf/homo_sapiens/Homo_sapiens.GRCh38.101.gtf.gz
gzip -d Homo_sapiens.GRCh38.101.gtf.gz

Build STAR Index with the following command:

STAR --runThreadN 16 --runMode genomeGenerate --genomeDir ./ \
     --genomeFastaFiles ./Homo_sapiens.GRCh38.dna.toplevel.fa \
     --sjdbGTFfile ./Homo_sapiens.GRCh38.101.gtf \
     --sjdbOverhang 100

2. Prepare Kraken2 database

wget ftp://ftp.ccb.jhu.edu/pub/data/kraken2_dbs/minikraken_8GB_202003.tgz
tar zxf minikraken_8GB_202003.tgz

How to use PathogenTrack?

Before running this tutorial, you should run cellranger or alevin to get the single cells' gene expression matrix. Here, we take the simulated 10X sequencing data as an example:

First, we use cellranger to get scRNA-seq expression matrix and valid barcodes:

cellranger count --id cellranger_out --transcriptom /path/to/cellranger_database/

Then we run PathogenTrack to identify and quantify pathogen expression at the single-cell level:

conda activate PathogenTrack
python PathogenTrack.py count --project_id PathogenTrack_out --pattern CCCCCCCCCCCCCCCCNNNNNNNNNN \
                              --min_reads 10 --confidence 0.11 --star_index ~/database/STAR_index/ \
                              --kraken_db ~/database/minikraken_8GB_20200312/ --barcode barcodes.tsv \
                              --read1 simulation_S1_L001_R1_001.fastq.gz \
                              --read2 simulation_S1_L001_R2_001.fastq.gz 

IMPORTANT: The Read 1 in the example is made up of 16 bp CB and 10 bp UMI, so the --pattern is CCCCCCCCCCCCCCCCNNNNNNNNNN (16C and 10N). Users must adjust the pattern with their own Read 1 accordingly.

Note: It may take 4-6 hours to complete one sample, and it depends on the performance of computational resources and the size of the raw single-cell data.

Please see QUICK_START.md for a complete tutorial.

Questions

For questions and suggestions about the pipeline or the code, please contact admin@ncrna.net and ty12260@rjh.com.cn. We will try our best to provide support, address new issues, and keep improving this software.