GMind

A novel plant UGTs annotated process for the mining of UGTs in plant genomes

Overview

GMind is a command line program (in Perl) based on Augustus, GlimmerHMM, which is used for sufficiently mining of UDP-glycosyltransferases from plant genomes. Firstly, the protein sequences of the known plant UGTs from an initial plant UGT database will be mapped into the un-annotated plant genomes, and the mapped regions will be extracted for the next annotation. Secondly, the extracted UGTs regions were independently de nove annotated by: 1) Augustus with the model training from all annotated UGTs in the 195 plant genomes, 2) Augustus with the model species of Arabidopsis thaliana and 3) GlimmerHMM with model species of A. thaliana. Thirdly, HMMER and UGT domain (PF00201) from Pfam were used to filter UGTs from all these annotated proteins. Fourthly, the filtered UGTs were evaluated by calculating accuracy score, the optimal UGTs were selected as candidates, while the UGTs with bad sores would be re-annotated by removing the good UGTs from the mapped regions. Finally, all these candidate UGTs would be output as the result of genomic annotation.

Installation

Running GMind requires a Linux system with Bash. The following dependencies need to be satisfied.

1 Install the following dependencies

Perl dependencies

Perl 5.10 or higher is required. Download the Core Package from http://www.bioperl.org

The following non-Core Perl modules are required:

• Bio::Perl
• List::Util

These modules are available at CPAN and can be installed/updated with

cpan Bio::Perl List::Util

HMMER

HMMER3 (available at http://hmmer.org/)

Add the path to environment:/your_path_to_HMMER3/bin

Augustus

Augustus3 (available at http://bioinf.uni-greifswald.de/augustus/binaries/)

Add the path to environment:/your_path_to_Augustus3/bin;

/your_path_to_Augustus3/scripts

AUGUSTUS_CONFIG_PATH=/your_path_to_Augustus3/config;

GlimmerHMM

GlimmerHMM (available at http://ccb.jhu.edu/software/glimmerhmm/)

Add the path to environment:/your_path_to_GlimmerHMM/bin;

export TrainPath=/your_path_to_GlimmerHMM/trained_dir/arabidopsis

Clustal Omega

Clustal Omega (available at http://www.clustal.org/omega/)

Add the path to environment:/your_path_to_Clustal_Omega/bin

Blast+

ncbi-blast+ (available at https://blast.ncbi.nlm.nih.gov/Blast.cgi?CMD=Web&PAGE_TYPE=BlastDocs&DOC_TYPE=Download)

Add the path to environment:/your_path_to_Blast+/bin

2 Download the Pfam

cd your_path_to_GMind
perl lib.pl

3 Add the trained plant UGT model to Augustus path

cp -r /your_path_to_GMind/lib/PUGT /your_path_to_Augustus/config/species/

4 Add the path to environment:

export GMind=/your_path_to_GMind/bin
export GMind_lib=/your_path_to_GMind/lib

Usage

Input

• Prepare the Genomic sequence file(s) and put it/them into a folder.
• Genomic sequence from the target species in FASTA format and named with its specie name linked with "_". (For example Arabidopsis_thaliana.fna, Zea_mays.fasta)

The tool is applicable to complete as well as draft genome assemblies. Every sequence needs to have a unique ID (first word of a FASTA header is used for ID). Examples of valid FASTA headers:

>contig10
sequence(ATGCT...)
> seq3  genome Z
sequence((ATGCT...))
>IV contig 25
sequence((ATGCT...))

Running GMind

To run GMind, use the following command:

perl $GMind/GMind.pl your_absolutepath_to_genomic_Folder CPU

Output

• The annotated proteins: your_path_to_genomic_Folder/Resultdir/GMind.pep
• The divergence score of each protein compared with PUGTDB: your_path_to_genomic_Folder/Resultdir/GMind.txt

About

Reference: [ Liu Yuqian, Wang Qian, Liu Xiaonan, Cheng Jian, et al. (2023). pUGTdb: A comprehensive database of plant UDP-dependent glycosyltransferases. Molecular Plant https://doi.org/10.1016/j.molp.2023.01.003. ]

Tianjin Institute of Industrial Biotechnology, Chinese Academy of Sciences, China.

Contact: jianglab_admin@tib.cas.cn.