shulp2211 / slideseq-tools

Tools for analyzing Slide-seq data

Slide-seq tools

Tools for analyzing Slide-seq data including building genome reference, aligning reads to reference genome, generating feature-barcode matrices, performing gene expression analysis and matching data from in situ sequencing and indexing of barcodes with short read sequencing data.

Requirement

Several public tools need to be pre-installed:

1. Drop-seq tools: https://github.com/broadinstitute/Drop-seq
2. Picard: https://broadinstitute.github.io/picard/
3. STAR: https://github.com/alexdobin/STAR
4. Java
5. Samtools
6. gcc/g++
7. Python (prefer 3.6 or above)

Several Python packages need to be installed for calculation and ploting:

1. numpy
2. pandas
3. plotnine
4. matplotlib

Build genome reference

The Slide-seq tools need a genome reference in specific format for alignment and analysis. You could build a genome reference based on input fasta and gtf files.

Command:

build_reference.py manifest_file

Check example.buildreference.txt for manifest file format

Run the Slide-seq tools

Add below commands into run.sh and build_reference.sh or your bashrc file (command might be different on your system):

use Java-1.8
use .samtools-1.7
use Python-3.6

Compile CMatcher (command might be different on your system):

g++ -std=c++11 -o cmatcher cmatcher.cpp

Submit a request to the Slide-seq tools:

python submit_job.py manifest_file

Notice:

1. Check example.manifest.txt for manifest file format
2. An email from [e-mail] will be sent to you if email_address is specified in the manifest file when the submission is received, the workflow finishes, and/or any job fails.
3. In order to speed up the process of NovaSeq data and NovaSeq S4 data, the Slide-seq tools split each lane into a few slices, run the alignment steps on the slices parallelly and combine the alignment outputs together.
4. See user_doc.txt for detailed usage of the Slide-seq tools.