# EEISP
EEISP identifies gene pairs that are codependent and mutually exclusive from single-cell RNA-seq data.
## Installation
pip3 install eeisp
### (Optional) Dependencies for GPU
EEISP requires [cupy](https://cupy.dev/) when using GPU computation `--gpu`. Use pip to install cupy like this (see [the manual](https://docs.cupy.dev/en/stable/install.html) for more detail).
# For CUDA 9.2
pip3 install cupy-cuda92
# For CUDA 10.1
pip3 install cupy-cuda101
If you do not use `--gpu`, you do not need to install cupy.
## Usage
EEISP takes a read count matrix as an input, in which rows and columns represent genes and cells, respectively. A gzipped file (.gz) is also acceptable.
0. (Optional) Convert CellRanger output to an input matrix (require R and [Seurat](https://satijalab.org/seurat/) library)
```
datadir="outs/filtered_feature_bc_matrix/"
matrix="matrix.txt"
R -e "library(Seurat); so <- Read10X('$datadir'); write.table(so, '$matrix', quote=F, sep=',', col.names=T)"
```
1. `eeisp` calculates the CDI and EEI scores for all gene pairs. The output contains lists of gene pairs that have CDI or EEI values above the specified threshold and the tables of degree distribution.
```
usage: eeisp [-h] [--threCDI THRECDI] [--threEEI THREEEI] [--tsv] [--gpu] [-p THREADS] [-v] matrix output
positional arguments:
matrix Input matrix
output Output prefix
optional arguments:
-h, --help show this help message and exit
--threCDI THRECDI Threshold for CDI (default: 20.0)
--threEEI THREEEI Threshold for EEI (default: 10.0)
--tsv Specify when the input file is tab-delimited (.tsv)
--gpu GPU mode
-p THREADS, --threads THREADS number of threads (default: 2)
-v, --version show program's version number and exit
```
2. `eeisp_add_genename_from_geneid` add Gene Names (Symbols) to the output files of `eeisp`.
```
usage: eeisp_add_genename_from_geneid [-h] [--i_id I_ID] [--i_name I_NAME] input output genelist
positional arguments:
input Input matrix
output Output prefix
genelist Gene list
optional arguments:
-h, --help show this help message and exit
--i_id I_ID column number of gene id (default: 0)
--i_name I_NAME column number of gene name (default: 1)
```
## Tutorial
The sample data is included in `sample` directory.
* `data.txt`: the input matrix of scRNA-seq data.
* `genelidlist.txt`: the gene list for `eeisp_add_genename_from_geneid`.
eeisp data.txt Sample --threCDI 0.5 --threEEI 0.5 -p 8
This command outputs gene pair lists that have CDI>0.5 or EEI>0.5. `-p 8` means 8 CPUs are used.
Supply `--gpu` option to GPU computation (require [cupy](https://www.preferred.jp/en/projects/cupy/)):
eeisp data.txt Sample --threCDI 0.5 --threEEI 0.5 -p 8 --gpu
(Note: Since GPU computation covers a part of eeisp, it is better to use multiple CPUs even in `--gpu` mode for the fast computation.)
Output files are:
```
Sample_CDI_score_data_thre0.5.txt # A list of gene pairs with CDI score.
Sample_CDI_degree_distribution_thre0.5.csv # A table of the number of CDI degree and genes.
Sample_EEI_score_data_thre0.5.txt # A list of gene pairs with EEI scores.
Sample_EEI_degree_distribution_thre0.5.csv # A table of the number of EEI degree and genes.
```
The output files might include gene ids only.
```
$ head Sample_CDI_score_data_thre0.5.txt
2 7 ESG000003 ESG000008 0.96384320244841
0 1 ESG000001 ESG000002 0.6852891560232545
0 6 ESG000001 ESG000007 0.6852891560232545
7 8 ESG000008 ESG000009 0.6852891560232545
3 9 ESG000004 ESG000010 0.6469554204484568
4 6 ESG100005 ESG000007 0.5258703930217091
```
If you want to add gene names (Symbols), use `eeisp_add_genename_from_geneid` with `geneidlist.txt`, which contains the pairs of gene ids and names.
```
eeisp_add_genename_from_geneid \
Sample_CDI_score_data_thre0.5.txt \
Sample_CDI_score_data_thre0.5.addgenename.txt \
geneidlist.txt
eeisp_add_genename_from_geneid \
Sample_EEI_score_data_thre0.5.txt \
Sample_EEI_score_data_thre0.5.addgenename.txt \
geneidlist.txt
```
The output files include gene names.
```
$ head Sample_CDI_score_data_thre0.5.addgenename.txt
2 7 ESG000003 ESG000008 OR4F5 FO538757.3 0.96384320244841
0 1 ESG000001 ESG000002 RP11-34P13.3 FAM138A 0.6852891560232545
0 6 ESG000001 ESG000007 RP11-34P13.3 RP11-34P13.9 0.6852891560232545
7 8 ESG000008 ESG000009 FO538757.3 FO538757.2 0.6852891560232545
3 9 ESG000004 ESG000010 RP11-34P13.7 AP006222.2 0.6469554204484568
4 6 ESG100005 ESG000007 RP11-34P13.8 RP11-34P13.9 0.5258703930217091
```
## Reference
Nakajima N., Hayashi T., Fujiki K., Shirahige K., Akiyama T., Akutsu T. and Nakato R., [Codependency and mutual exclusivity for gene community detection from sparse single-cell transcriptome data](https://academic.oup.com/nar/advance-article/doi/10.1093/nar/gkab601/6324613), *Nucleic Acids Research*, 2021.
Raw data
{
"_id": null,
"home_page": "https://github.com/nakatolab/EEISP",
"name": "eeisp",
"maintainer": "",
"docs_url": null,
"requires_python": ">=3.6",
"maintainer_email": "",
"keywords": "eeisp scRNA-seq",
"author": "Natsu Nakajima, Ryuichiro Nakato",
"author_email": "rnakato@iqb.u-tokyo.ac.jp",
"download_url": "",
"platform": null,
"description": "# EEISP\n\nEEISP identifies gene pairs that are codependent and mutually exclusive from single-cell RNA-seq data. \n \n## Installation\n\n pip3 install eeisp\n\n### (Optional) Dependencies for GPU\nEEISP requires [cupy](https://cupy.dev/) when using GPU computation `--gpu`. Use pip to install cupy like this (see [the manual](https://docs.cupy.dev/en/stable/install.html) for more detail).\n\n # For CUDA 9.2\n pip3 install cupy-cuda92\n # For CUDA 10.1\n pip3 install cupy-cuda101\n\nIf you do not use `--gpu`, you do not need to install cupy.\n\n## Usage\nEEISP takes a read count matrix as an input, in which rows and columns represent genes and cells, respectively. A gzipped file (.gz) is also acceptable.\n\n 0. (Optional) Convert CellRanger output to an input matrix (require R and [Seurat](https://satijalab.org/seurat/) library)\n ```\n datadir=\"outs/filtered_feature_bc_matrix/\"\n matrix=\"matrix.txt\"\n R -e \"library(Seurat); so <- Read10X('$datadir'); write.table(so, '$matrix', quote=F, sep=',', col.names=T)\"\n ```\n\n 1. `eeisp` calculates the CDI and EEI scores for all gene pairs. The output contains lists of gene pairs that have CDI or EEI values above the specified threshold and the tables of degree distribution.\n ```\n usage: eeisp [-h] [--threCDI THRECDI] [--threEEI THREEEI] [--tsv] [--gpu] [-p THREADS] [-v] matrix output\n\n positional arguments:\n matrix Input matrix\n output Output prefix\n\n optional arguments:\n -h, --help show this help message and exit\n --threCDI THRECDI Threshold for CDI (default: 20.0)\n --threEEI THREEEI Threshold for EEI (default: 10.0)\n --tsv Specify when the input file is tab-delimited (.tsv)\n --gpu GPU mode\n -p THREADS, --threads THREADS number of threads (default: 2)\n -v, --version show program's version number and exit\n ``` \n 2. `eeisp_add_genename_from_geneid` add Gene Names (Symbols) to the output files of `eeisp`.\n ```\n usage: eeisp_add_genename_from_geneid [-h] [--i_id I_ID] [--i_name I_NAME] input output genelist\n\n positional arguments:\n input Input matrix\n output Output prefix\n genelist Gene list\n\n optional arguments:\n -h, --help show this help message and exit\n --i_id I_ID column number of gene id (default: 0)\n --i_name I_NAME column number of gene name (default: 1)\n ```\n## Tutorial\nThe sample data is included in `sample` directory. \n * `data.txt`: the input matrix of scRNA-seq data.\n * `genelidlist.txt`: the gene list for `eeisp_add_genename_from_geneid`.\n\n\n eeisp data.txt Sample --threCDI 0.5 --threEEI 0.5 -p 8\nThis command outputs gene pair lists that have CDI>0.5 or EEI>0.5. `-p 8` means 8 CPUs are used.\n\nSupply `--gpu` option to GPU computation (require [cupy](https://www.preferred.jp/en/projects/cupy/)):\n\n eeisp data.txt Sample --threCDI 0.5 --threEEI 0.5 -p 8 --gpu\n \n(Note: Since GPU computation covers a part of eeisp, it is better to use multiple CPUs even in `--gpu` mode for the fast computation.)\n\nOutput files are:\n```\n Sample_CDI_score_data_thre0.5.txt # A list of gene pairs with CDI score. \n Sample_CDI_degree_distribution_thre0.5.csv # A table of the number of CDI degree and genes. \n Sample_EEI_score_data_thre0.5.txt # A list of gene pairs with EEI scores. \n Sample_EEI_degree_distribution_thre0.5.csv # A table of the number of EEI degree and genes.\n```\nThe output files might include gene ids only. \n\n```\n $ head Sample_CDI_score_data_thre0.5.txt\n 2 7 ESG000003 ESG000008 0.96384320244841\n 0 1 ESG000001 ESG000002 0.6852891560232545\n 0 6 ESG000001 ESG000007 0.6852891560232545\n 7 8 ESG000008 ESG000009 0.6852891560232545\n 3 9 ESG000004 ESG000010 0.6469554204484568\n 4 6 ESG100005 ESG000007 0.5258703930217091\n```\n\nIf you want to add gene names (Symbols), use `eeisp_add_genename_from_geneid` with `geneidlist.txt`, which contains the pairs of gene ids and names.\n\n```\n eeisp_add_genename_from_geneid \\\n Sample_CDI_score_data_thre0.5.txt \\\n Sample_CDI_score_data_thre0.5.addgenename.txt \\\n geneidlist.txt\n eeisp_add_genename_from_geneid \\\n Sample_EEI_score_data_thre0.5.txt \\\n Sample_EEI_score_data_thre0.5.addgenename.txt \\\n geneidlist.txt\n```\n\nThe output files include gene names.\n\n```\n $ head Sample_CDI_score_data_thre0.5.addgenename.txt\n 2 7 ESG000003 ESG000008 OR4F5 FO538757.3 0.96384320244841\n 0 1 ESG000001 ESG000002 RP11-34P13.3 FAM138A 0.6852891560232545\n 0 6 ESG000001 ESG000007 RP11-34P13.3 RP11-34P13.9 0.6852891560232545\n 7 8 ESG000008 ESG000009 FO538757.3 FO538757.2 0.6852891560232545\n 3 9 ESG000004 ESG000010 RP11-34P13.7 AP006222.2 0.6469554204484568\n 4 6 ESG100005 ESG000007 RP11-34P13.8 RP11-34P13.9 0.5258703930217091\n```\n\n## Reference\nNakajima N., Hayashi T., Fujiki K., Shirahige K., Akiyama T., Akutsu T. and Nakato R., [Codependency and mutual exclusivity for gene community detection from sparse single-cell transcriptome data](https://academic.oup.com/nar/advance-article/doi/10.1093/nar/gkab601/6324613), *Nucleic Acids Research*, 2021.\n",
"bugtrack_url": null,
"license": "GPL3.0",
"summary": "Identify gene pairs that are codependent and mutually exclusive from single-cell RNA-seq data.",
"version": "0.5.0",
"split_keywords": [
"eeisp",
"scrna-seq"
],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "3be3c31759115a7c6d0bba728bdf5ff15f55de4247fc0bdc5fb99ef58905b9f1",
"md5": "497524125165da6b16d057a268b7e5e1",
"sha256": "1524a84f50f8fddb8206f79fa3cdd23dd91d91c18d8ff74e57df023ab866ea6b"
},
"downloads": -1,
"filename": "eeisp-0.5.0-py3-none-any.whl",
"has_sig": false,
"md5_digest": "497524125165da6b16d057a268b7e5e1",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.6",
"size": 19370,
"upload_time": "2023-04-12T06:10:21",
"upload_time_iso_8601": "2023-04-12T06:10:21.801816Z",
"url": "https://files.pythonhosted.org/packages/3b/e3/c31759115a7c6d0bba728bdf5ff15f55de4247fc0bdc5fb99ef58905b9f1/eeisp-0.5.0-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2023-04-12 06:10:21",
"github": true,
"gitlab": false,
"bitbucket": false,
"github_user": "nakatolab",
"github_project": "EEISP",
"travis_ci": false,
"coveralls": false,
"github_actions": false,
"lcname": "eeisp"
}
JSON
