Introduction

Welcome to the full-length Homology-based R-gene Prediction (HRP) project. This project aims to develop an informatics pipeline for genome-wide resistance (R)-genes (also known as NB-LRR genes) prediction and classification. We sincerely wish our method can significantly benefit your R-gene study, and enjoy high-performances.

Table of Contents to use HRP

• Essential software
• Example commands
  • 1. Data preparation
  • 2. Identification of NB domain using a protein domain search (PDS)
  • 3. Decomposition into motifs of NB Pfam domains
  • 4. Identification of additional NB Pfam domain using motif search
  • 5. Identification of LRR domain using a protein domain search (PDS)
  • 6. Classification in full-length or partial NB-LRR genes
  • 7. de novo prediction of NB-LRR genes
  • 8. Selection of non-redundant NB-LRR genes
  • 9. Annotation of NB-LRR proteins
  • 10. Comparison and visual inspection of NB-LRR genes predicted by HRP
• Bash script
  • IPS2fpGs.sh

Essential software

Make sure below programs are correctly installed according to their installation manual.

• InterProScan, a HMM based domain identification package (Jones et al., 2014)

• MEME Suite, command line programs to discover novel and ungapped motifs (Bailey et al., 2006)

• BEDTools utilities, tools for a wide-range of genomics analysis tasks (Quinlan and Hall, 2010)

• genBlastG, algorithm that uses the HSPs reported by BLAST to define high-quality gene models (She et al., 2011)

• AGAT Suite, tools to handle gene annotations in any GTF/GFF format (Dainat, 2020)

• IGV, interactive tool for the visual exploration of genomic data (Thorvaldsdóttir et al., 2013)

Example commands

Please follow the example commands for execution instruction of HRP method

1. Data preparation

Download the genome sequence and the protein sequences (encoded by gene set) of interest, for example related to the sugar beet genome assembly RefBeet-1.2

wget -O RefBeet-1-2_proteins.fasta https://bvseq.boku.ac.at/Genome/Download/RefBeet-1.2/BeetSet-2.genes.1408.pep
wget -O RefBeet-1-2_genome.fasta https://bvseq.boku.ac.at/Genome/Download/RefBeet-1.2/RefBeet-1.2.fna.gz

2. Identification of NB domain using a protein domain search (PDS)

An InterProScan example command using the data from 1

interproscan -f TSV -app Pfam -i RefBeet-1-2_proteins.fasta -b RefBeet-1-2_proteins	

3. Decomposition into motifs of NB Pfam domains

3.1. BED file formatting of NB Pfam domains using the data from 2

grep NB-ARC RefBeet-1-2_proteins.tsv | cut -f1,7,8 > NB.bed

3.2. Extraction of NB Pfam domain sequences using the data from 1 and 3.1

bedtools getfasta -fi RefBeet-1-2_proteins.fasta -bed NB.bed -fo NB_Pfam_Domain_Sequences.fasta

3.3. A MEME example command using data from 3.2

meme NB_Pfam_Domain_Sequences.fasta -protein -o meme_out -protein -mod zoops -nmotifs 19 -minw 4 -maxw 7 -objfun classic -markov_order 0

4. Identification of additional NB Pfam domain using motif search

A MAST example command using data from 1 and 3.3

mast -o mast_out meme_out.txt RefBeet-1-2_proteins.fasta

5. Identification of LRR domain using a protein domain search (PDS)

An InterProScan example command using sequences from 1, filtered on the base of gene-ID list from 2 and 4

interproscan -f TSV -app SUPERFAMILY -i RefBeet-1-2_proteins_subset.fasta -b RefBeet-1-2_proteins	

6. Classification in full-length or partial NB-LRR genes

IPS2fpGs.sh command example using data from 2 and 5

IPS2FPGs RefBeet-1-2_proteins.tsv -o output_file.tsv

7. de novo prediction of NB-LRR genes

A genBlastG example command using data from 6

genblastG -q Full-length_NB-LRRs.fasta -t RefBeet-1-2_genome.fasta -gff -cdna -pro -o genblastG-output

8. Selection of non-redundant NB-LRR genes

8.1. Filtering of NB-LRR gene models longer than 20 kbp using the data from 7

agat_sp_filter_gene_by_length.pl --gff genblastG-output.gff --size 20000 --test "<" -o genblastG-output_FbL.gff

8.2. Identification of overlapped gene models using the data from 8.1

grep transcript genblastG-output_FbL.gff | gff2bed | sortBed | clusterBed -s | cut -f4,11  > genblastG-output_FbL_clusters

8.3. Estimation of NB-LRR protein sequence lengths using the data from 7

awk 'BEGIN{FS="[> ]"} /^>/{val=$2;next}  {print val,length($0);val=""} END{if(val!=""){print val}}' genblastG-output.pro | tr ' ' \\t > genblastG-output_FbL_length

8.4. List of non-redundant NB-LRR gene models using the data from 8.2 and 8.3

join -t $'\t' -1 1 -2 1 -o 1.1,1.2,2.2 <( sort -bk1 genblastG-output_FbL_clusters) <(sort -bk1 genblastG-output_FbL_length) | sort -bk2,2 -bk3,3 -nr | sort -uk2,2 | cut -f1 > R-gene_ID_list

9. Annotation of NB-LRR proteins

An InterProScan example command using the data extracted from 6 on the base of 8.4

interproscan -f TSV,GFF3 -i NB-LRR_gene_candidates.fasta -b RefBeet-1-2_B-LRR_gene_candidates

10. Comparison and visual inspection of NB-LRR genes predicted by HRP using data from 8.1 and 9

The candidate R loci identified from HRP could be exported in GFF format and imported into the IGV genome browser for comparison and visual inspection

Bash script

IPS2fpGs.sh,

The aim of this script is to feed it InterProScan results and return the full-length and partial NB-LRR genes classification on the base of R protein domains. Users can download and use the IPS2FPGs freely for research purpose only with acknowledgment to us and quoting HRP paper as reference