Update notes

May 20th, 2024

(1) Fix bugs to meet the gtf format generate by StringTie;

(2) Improve alignment scoring system

December 18th, 2022

Fix bugs

September 23rd, 2022

(1) optimize some functions;

(2) add a visualization tool to plot EGI and isoforms of two species.

June 6th, 2022

Increase the running speed.

Introduction

EGIO (Exon Group Ideogram based detection of Orthologous exons and Orthologous isoforms) is aimed to detect orthologous exons and isoforms within pre-defined orthologous gene pairs. EGIO uses a dynamic programming strategy to do isoform alignment, in which reciprocal BLASTN results are used to guide the whole process.

The scripts have been tested on MacOS (X64) and Ubuntu (Linux).

Requirement

(1) gcc

gcc is required to compile ___pairwisealign.c

(2) EGIO required python packages pandas and numpy, to install pandas and numpy:

pip install pandas

pip install numpy

(3) To increase the accuracy, the algorithm uses a BLASTN guided model, so a reciprocal BLASTN of exons is used. To run reciprocal BLASTN, BLAST+ is required before running EGIO, which could be downloaded at https://ftp.ncbi.nlm.nih.gov/blast/executables/blast+/LATEST/. For linux, it is easy to install BLAST+ using:

apt-get install ncbi-blast+

(4) A tab seperated file containing pre-defined orthologous gene pairs, which could be prepared according to Inparanoid: https://inparanoid.sbc.su.se/cgi-bin/index.cgi. The gene id provided by Inparanoid is Uniprot ID, which can be transformed to Ensembl Gene ID using gene annotations in Ensembl BioMart. An example is listed here:

hsa ptr

ENSG00000131018 ENSPTRG00000018717

ENSG00000183091 ENSPTRG00000012536

ENSG00000183091 ENSPTRG00000012536

... 

To be noted, the header is required in the file.

(4) Required files, which can be downloaded from Ensembl (http://asia.ensembl.org/info/data/ftp/index.html):

cDNA fasta files,

CDS fasta files,

reference transcriptome gtf files

(5) plotnine package is required, if you want to visuzalize isoforms of two orthologous genes

pip install plotnine

Usage

Run EGIO

To test EGIO, please download example files in https://github.com/wu-lab-egio/EGIO_example_source, and unzip gtf files. Put these files in folder "example", and put the "example" folder into "EGIO-main". When running EGIO, it may notice that it is not permitted to run _Run_egio.sh, and it can be solved by chmod command.

To run EGIO, type following commands in Ternimal:

cd /path/to/EGIO-main
chmod 777 _RUN_egio.sh
./_RUN_egio.sh -s hsa -S ptr -r example/hsa.gtf -R example/ptr.gtf -e example/hsa_mRNA_example.fa -E example/ptr_mRNA_example.fa -o example/hsa_CDS_example.fa -O example/ptr_CDS_example.fa -h example/homogene.txt -p 6 -i 0.8 -c 0.8 -m 2 -n -2 -g -1

It will take about 1.5 hours to complete the comparison of human and chimpanzee using 6 cores (MacOS, 2.4 GHz, 16 RAM).

Visualization

plotnine package is required to visualize isoforms of two orthologous genes:

pip install plotnine

Visualization:

python __plotIsoCom.py --speciespair species1-species2 --orthologpair gene_of_species1-gene_of_species2 --feature cds_or_exon --extrapath /path/to/extrainfo/folder --egpath /path/to/EGfile --outpath /path/to/store/plot

Parameters:

-s: name of species1, to be noted, the name should be consistent with the header of the file containing pre-defined orthologous gene pairs, eg: hsa (required)

-S: name of species2, to be noted, the name should be consistent with the header of the file containing pre-defined orthologous gene pairs, eg: ptr (required)

-r: gtf file of species1, reference transcriptome (required)

-R: gtf file of species2, reference transcriptome (required)

-e: cDNA fasta file of species1, sequence of total exons (required)

-E: cDNA fasta file of species2, sequence of total exons (required)

-o: CDS fasta file of species1, sequence of orf (required)

-O: CDS fasta file of species2, sequence of orf (required)

-h: orthologous/homologous gene pair containing file (required)

-p: multiple processing to run EGIO (optional, default is 1)

-i: identity threshold to detect orthologous exons during dynamic programming (optional, default is 0.8)

-c: coverage threshold to detect orthologous exons during dynamic programming (optional, default is 0.8)

-m: match score during dynamic programming (optional, default is 2)

-n: mismatch penalty during dynamic programming (optional, default is -2)

-g: gap penalty during dynamic programming (optional, default is -1)

Description of EGIO results

Two files will be generated after running EGIO: ExonGroup.txt and OrthoIso.txt.

Explaination of ExonGroup.txt headers:

Group: exon group number

species1EnsemblG: gene id of species1

species1Pos: unique exon region of a exon group in species1

species2EnsemblG: gene id of species2

species2Pos: unique exon region of a exon group in species2

Iden: identity of unique exon region in two species of a exon group

Type: type of exon group.

Explaination of OrthoIso.txt headers:

species1: gene id of species1

species2: gene id of species1

species1iso: isoform of species1

species2iso: isoform of species2

exoniden: identity of corresponding exons. The exon IDs can be found in corresponding species.exon file, which is one of the EGIO output in the "extrainfo" folder in the work directory.

Reference:

Ma J, Wu JY, Zhu L. Detection of orthologous exons and isoforms using EGIO. Bioinformatics. 2022 10;38(19):4474-4480.