spacemake 0.0.3

A bioinformatic pipeline for the analysis of spatial transcriptomic data

Spatial transcriptomics sequencing

Structure of the pipeline

This repository collects all scripts and tools used for analyzing the sequencing side of the spatial transcriptomics datasets. The following steps are currently performed:

Demultiplex the data

This assumes that the sample sheet has been provided and that the raw data has been copied to the basecalls folder. The tool bcl2fastq is used to demultiplex the data.

Extract Bead Barcode and UMI from R1 and R2

Depending on the R1 and R2 structure, which can be defined in the top-level config.yaml, spacemake will create a tagged but unmapped.bam, with CB (Bead Barcode) and MI (UMI) tags. This .bam file will then be used as an input by STAR

Processing

Once we extracted the CB and MI tags, the data is processed as follows:

• Adapter removal (R2)
• PolyA removal (R2)
• Map reads with STAR (species can be defined in either the samplesheet or in the config.yaml)
• Postprocessing
• Join data with the bead position information
• Analyse the data automatically, produce reports

Inputs

Samples

Each sample can be either defined in the projects variable in the config.yaml or in the additional_projects. The first expects a list of elements with each having two variables: flowcell_id and sample_sheet. The second expects several variables: project_id, sample_id, puck_id, species, R1 and R2

Configuration

The pipeline is implemented in snakemake. All metadata of the experiments (experiment_name, flowcell_id, species, etc) should be put in a config.yaml file. An example config.yaml file is in the root of this repo.

To run the snakemake script, the snakemake python library is required (installed with pip or conda). The script requires at least 6 threads to run, this is due to pipeing several commands one after the other to descrease runtime.

Example run:

snakemake --snakefile path_to_snakefile --configfile path_to_configfile.

This will create the output in the directory in which the command is run. Note, that all samplesheet-flowcell_id paris should be ideally in one configfile somewhere.

Produced directory structure

The following directory structure will be produced by the snakemake file

    .
    |-- demultiplex_data                            # demultiplexed data folders, one per each samplesheet
    |   |-- 200110_STS_017_4-7-8STS_018_1           # directory names are identical to the samplesheet names
    |   |   |-- Stats
    |   |   |-- sts_017
    |   |   |-- sts_018
    |   |   |-- Undetermined_S0_R1_001.fastq.gz
    |   |   `-- Undetermined_S0_R2_001.fastq.gz
    |   `-- 20191206_spatseq_smples3-4
    |       |-- indicator.log
    |       |-- Reports
    |       |-- Stats
    |       |-- sts_0xxx
    |       |-- Undetermined_S0_R1_001.fastq.gz
    |       `-- Undetermined_S0_R2_001.fastq.gz
    |-- sts_017                                     # root output directory, one per project
    |   |-- data
    |   |   |-- sts_017_4                           # directory containing results of running the pipeline. one per sample 
    |   |   |-- sts_017_7
    |   |   `-- sts_017_8
    |   `-- reads                                   # reads directory, one per sample
    |       |-- fastqc
    |       |-- raw
    |       `-- reversed
    |-- sts_018
    |   |-- data
    |   |   `-- sts_018_1
    |   `-- reads
    |       |-- fastqc
    |       |-- raw
    |       `-- reversed
    `-- sts_0xxx
        |-- data
        |   |-- sts_01
        |   `-- sts_02
        `-- reads
            |-- fastqc
            |-- raw
            `-- reversed