https://github.com/ggonnella/subclinical_dengue_scrnaseq
https://github.com/ggonnella/subclinical_dengue_scrnaseq
Last synced: about 2 months ago
JSON representation
- Host: GitHub
- URL: https://github.com/ggonnella/subclinical_dengue_scrnaseq
- Owner: ggonnella
- Created: 2025-01-24T10:44:11.000Z (4 months ago)
- Default Branch: main
- Last Pushed: 2025-01-29T01:56:10.000Z (4 months ago)
- Last Synced: 2025-01-29T02:28:18.464Z (4 months ago)
- Language: R
- Size: 306 KB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Requirements
The analyses are perfomed using:
- R version 4.4.2
- 10X Cellranger v6 or v7
- 10X enclone v0.5.219
- mixcr v3.0.12
- Seurat v5, Platypus and several further R libraries (see ``R_LIBRARIES`` file)
# Organization
The repository is organized as follows:
- "cellranger" are the scripts/metadata used for running Cellranger
- "analysis-1" processing and visualization steps and results
- "helpers" helper scripts, containing functions common to multiple steps
- "scripts" command line scripts (e.g. to run all steps)
## Processing and Visualization Steps
Each step under "analysis-N" is classified as a "processing step" or
"visualization step".
Thereby:
- processing steps take the previous step data and generate
new further processed data (e.g. filtering, normalization, clustering, etc.);
for most steps (R), the data is saved in the ``rds`` format under
``results/vars``
- visualization steps generate plots (in ``results/plots``) and tables
(in ``results/tables``) but do not change the underlying data structure.
### Steps naming
Each step is named as follows
``stepnumber.versionnumber.type.celltype.description``
matching the following regular expression:
``(\d\d\d)\.(\d)\.([PV])\.([BT])\.(.*)``
where:
- stepnumber is the incremental number of the step,
starting from 001 for the T cells and 101 for the B cells analysis
- versionnumber is a version of the step, if the step is revised
- type is either P for processing or V for visualization
- celltype is either B for B cells or T for T cells
- description is a short textual description of the step
### Step organization
Each step is organized as follows:
- ``scripts`` contains the scripts (in most cases, R markdown) for the step
- ``results`` contains the results of the step
- ``rundir`` contains the output of the step (e.g. logs, intermediate files)
The scripts under scripts are symlinked into rundir and run there. The scripts
create the results directory and store the results there.
# Running the analysis
## From reads to counts: Cellranger
To run Cellranger, change the ``cellranger/metadata`` files to match the paths
to the fastq files and the reference genome. Then run the
``cellranger/scripts/run_all.sh`` script.
## Running the R analysis pipeline
The R analysis pipeline is organized in steps, which can be run individually
or in sequence. The input to the pipeline is the output of Cellranger, which
is stored in the ``cellranger/results`` directory.
### Running from Cellranger counts
To run the pipeline from the Cellranger results, instead of running Cellrager,
these must be put in the
``cellranger/results`` directory
and the correct directory structure must be re-created. Thereby .h5 files must
be put in directories under results named
``sample_(n)_[TB]/per_sample_outs/sample_1_[TB]/count/`` and the .csv files
must be put in directories under results named
``sample_(n)_[TB]/multi/vdj_[TB]/``.
### Upstream analysis (individual samples)
To run the upstream analysis, one case use the ``run_all_samples.sh`` scripts
in the 001/002 (T cells) and 101/102 (B cells) P and V steps.
The suggested way to do this is to create links to all the content of scripts
under a new directory called rundir, and run the scripts from there.
Thereby it is necessary to specify the own paths to this repository
(prjpath parameter) and to R library containing Seurat 5 (libpath).
### Downstream analysis (merged samples)
To run the Rmd files, one can use the knit2html script under scripts
in the steps from 003/103 onwards.
The suggested way to do this is to create links to all the content of scripts
under a new directory called rundir, and run the scripts from there.
When running the scripts it is necessary to specify the own paths to this
repository (prjpath parameter) and to R library containing Seurat 5 (libpath).
### Running all steps (including secondary steps)
The ``run_[TB]_cell_analysis.sh`` script run all the steps for the
T and B analysis starting from the Cellranger output.
Thereby it is necessary to specify the own paths to this repository
(prjpath parameter) and to R library containing Seurat 5 (libpath).
This script runs all the processing and visualization steps, including
secondary steps which were not used for the manuscript figures.
### Running steps for manuscript figures only
The steps necessary to generate the figures for the manuscript can be run
using the ``run_minimal_[BT]_from_counts.sh`` scripts, located in the
steps directory.
### Running downstream analyses only (after Harmony integration)
The downstream analyses necessary for to generate the figures can be run by
storing the integrated data in a directory called ``integrated/T`` and
``integrated/B`` respectively, and running all steps from that script onwards,
e.g. using the scripts in the ``steps`` directory
``run_minimal_[BT]_from_int.sh``, which only run the minimal set of steps
needed to generate the figures for the manuscript.
## Collecting figure elements
First run the pipeline, e.g. using the ``run_[TB]_cell_analysis.sh`` script
(run all steps, including non secondary analyses) or the
``run_minimal_[BT]_from_counts.sh`` script (run steps necessary to produce the
figures, from Cellranger results), or ``run_minimal_[BT]_from_int.sh``
(run steps necessary to produce the figures, from integrated data).
The results of the pipeline for the manuscript figures and supplementary
figures can be then collected after running the pipeline
using the ``collect_results.sh`` script under ``analysis-1/results``.