DNAlogy

Light bioinformatics tools for gene set statistical analysis and management based on DNA-seq (meta) data
 

• Tool 1 — fastGSEA
  • Installation
  • Usage
  • Examples
  • Workflow
  • Methods
  • Contributions
  • Main dependencies
• Tool 2 — Javascript API

• Licence
  • Author

 

Tool 1 — fastGSEA

FastGSEA (fast Gene Set Enrichment Analysis) performs GO-terms enrichment analysis bewteen two gene sets, based on hypergeometric tests. These gene sets must be provided as text files containing one international databank (ncbi, refseq, etc) gene or proteins identifier per line. FastGSEA can also be used only as a standalone mapping tool, using the --mapOnly option.

In every cases of use, supported ids are:

1. UniProtKB-AC
2. RefSeq
3. UniProtKB-ID
4. GeneID (EntrezGene)
5. GI
6. GO (as output only!)
7. UniRef100
8. UniRef90
9. UniRef50
10. UniParc
11. UniGene
12. EMBL
13. EMBL-CDS
14. Ensembl
15. Ensembl_TRS
16. Ensembl_PRO

---

Installation

Only on GNU/Linux based systems. You can install it using Conda. If you already have Conda on your system you can apply the following instructions:

Auto-install

Just go to the FAST_GSEA directory, then run install.sh

# Go to fastGSEA directory
cd DNAlogy/FAST_GSEA

# Run install script
bash install.sh 

# Respond "yes" to the "Do you wish the installer to prepend the Miniconda2 install location to PATH in your /home/username/.bashrc ?" answer.

Next, you'll just have to activate the gsea_env environment when you want to use fastGSEA.

source activate gsea_env

Then, type fastGSEA and you will able to use it! If fastGSEA command is not recognized, you may need to reload your shell environnement by typing source ~/.bashrc.

Manual install

Use the packages.yml to retrieve all Python and R dependencies then install fastGSEA:

# Go to fastGSEA directory
cd DNAlogy/FAST_GSEA

# Create the environment from the yaml file
conda env create -f packages.yml 

# Activate the enrivonment:
source activate gsea_env

# Add fastGSEA to your environment variables
PATH="src/:${PATH}"
export PATH

# or add an alias for fastGSEA in your bashrc
echo 'alias fastGSEA="python $(pwd)'/src/fastGSEA.py'"' >> ~/.bashrc

# Reload your shell (here, bash) settings 
source ~/.bashrc

---

Usage

Command line options

Option	Description	Required
-ech	Sample ids file (one id per line)	Yes
-univ	Universe ids file (one id per line)	Yes
-mappingFile	idmapping_selected.tab.gz file	Yes
-output	Output results prefix	Yes
-obo	Gene ontology .obo graph file used when "--trim" option activated	No
-toDB	databank identifier wanted as output when "--mapOnly" option activated	No
--fromOtherDB	Activate all ids support (slower)	No
--mapOffline	Perform "MAP" step offline	No
--trim	Trim prokaryotic GO-terms	No
--view	Outputs the enrichment results in a 2D graph	No
--mapOnly	Perform only the "MAP" step and keep its results	No
--keepTmp	Keep temporary files folder	No

Usage - Where can I find the -mappingFile and -obo files?

-mappingFile is used for offline ids mappping and can be found on the Uniprot FTP ( > 6gb file)

-obo is the gene ontology graph used for GO-terms checking and trimming. It is availaible on the Open Biological and Biomedical Ontology (OBO) subset file. There are daily releases, so you can download the latest ones here

⚠️ Important note: be careful when trimming non prokaryotic GO-terms, 'gosubset_prok' terms are not maitained since 2018/06 because some of them muight be irrelevant. More information here and here.

Usage — Input files format

FastGSEA takes two input files (one for sample, second one for universe). They have to be text files containing one international databank (ncbi, refseq, etc) supported identifiers (listed above) per line, for example:

O55719
Q6GZM9
Q6GZM8
NP_302218.1
WP_008262748.1
WP_011437797.1
NP_149806.1
NP_854636.1
WP_003877490.1
Q6GZM7
P0C9F0
P0C9F1
P0C9F2
P0C9E9
Q65209
P0C9F4
P0C9F5
P0C9F6

---

Examples

All data used for the following examples are avalaible in the examples directory.

Example — Id mapping: map any ids to UniRef100 ids

This command line requires two outputs, if you want to perform id mapping in only one file, just provide it twice as -ech and -univ.

# -ech: gene set sample ; -univ: gene set universe ; other args? please read the docs:)
fastGSEA -ech ech.txt  -univ univ.txt  -mappingFile idmapping_very_light.gz --mapOnly -toDB UniRef100 -output maybe/here

Examples — Gene set enrichment analysis: find which GO-terms from a gene set are overrepresented

With most steps offline (faster, the better updated your -mappingFile and/or -obo are, the better the results will be):

# -ech: gene set sample ; -univ: gene set universe ; other args? please read the docs:)
fastGSEA -ech ech.txt  -univ univ.txt  -mappingFile idmapping_very_light.gz  --mapOffline -output maybe/here

Requesting NCBI and Uniprot APIs (most reliable, but slower):

# -ech: gene set sample ; -univ: gene set universe ; other args? please read the docs:)
fastGSEA -ech ech.txt  -univ univ.txt  -mappingFile idmapping_very_light.gz  -output maybe/here

...plus trimming obsolete and non Prokaryotic GO-terms (up to date obo file gosubset_prok.obo needed):

# -ech: gene set sample ; -univ: gene set universe ; other args? please read the docs:)
fastGSEA -ech ech.txt  -univ univ.txt  -mappingFile idmapping_very_light.gz -obo gosubset_prok.obo -output somewhere --trim

...plus generating a chart for enriched GO-terms:

# -ech: gene set sample ; -univ: gene set universe ; other args? please read the docs:)
fastGSEA -ech ech.txt  -univ univ.txt  -mappingFile idmapping_very_light.gz -obo gosubset_prok.obo -output somewhere --trim --view

All these options can be combined to use FastGSEA as you like. Examples of results data are also provided here. For example, the top 5 of all the enriched terms detected in the dummy dataset:

GO:ID	Go term	Number of hits	Expected number of hits	Go level	P-value	Corrected p-value	Aspect
GO:0044068	modulation by symbiont of host cellular process	1	0.02053442	6	0.0001562	0.2967242	BP
GO:0016791	phosphatase activity	4	0.5206708	6	0.000157	0.0408521	MF
GO:0042578	phosphoric ester hydrolase activity	4	0.5206708	5	0.000157	0.0408521	MF
GO:0016788	hydrolase activity, acting on ester bonds	4	0.5553822	4	0.0002142	0.0557149	MF
GO:0044003	modification by symbiont of host morphology	1	0.02566803	5	0.0002595	0.4928665	BP

And its associated 2D chart ( GO-term level = f(log p-value) ):

---

Workflow

As said previously, workflow can be stopped at each step, the last three parts of the workflow are optional and and behave as you set it up for.

You can also import all or part of the map enrich trim modules for another usage in your own scripts.

---

Methods

FastGSEA comes with several methods that you can manipulate to make it behave as you like. Fore more details, please read the technical documentation.

---

Contributions

Submit problems or requests using the Issue Tracker.

Want to contribute? Opened to all suggestions and pull requests.

---

Main dependencies

FastGSEA was developped using python 2.7 (fully compatible with 2.7...3.7+) and R 3.3.2

• ggplot2 - Version 2.2.0
• GO.db - Version 3.4.0
• gridExtra - Version 2.2.1

Tool 2 — Javascript API

~ Q4 2018 (W.I.P)

Licence

All DNAlogy tools are under GPL v3 licence.

Author

• Kévin Durimel
• Website: http://kevin.durimel.fr/