LTRfind: a pipeline for get LTR and LAI ( base on LTR_retriever)

0.Require software (we provided version.only for Linux X86_64)

• LTR_FINDER_parallel
• gt
• seqkit
• LTR_retriever V2.9.6
• LTR_HARVEST_parallel
• R ggplot2 packages (for visualization only)

1. Install LTR_retriever by conda

conda create -n LTR_retriever
conda activate LTR_retriever
conda install -y -c conda-forge perl perl-text-soundex
conda install -y -c bioconda cd-hit repeatmasker
git clone https://github.com/oushujun/LTR_retriever.git
./LTR_retriever/LTR_retriever -h

chmod 757 gt-1.6.2/bin/gt
chmod 757 seqkit
chmod 757 LTRfind
chmod 757 LTR_FINDER_parallel-1.1/bin/LTR_FINDER.x86_64-1.0.7/ltr_finder


./LTRfind -h

you also can add the path of LTRfind to the ~/.bashrc.

2.how to use it

It is mainly used for the identification of diploid and polyploid LTR and the calculation of LAI value.

Modify config.ini at first

Purpose:

Get the LTR_RTs in the genome and calculate the LAI value of the genome at the same time. for Diploid use (-D|Diploid) for Polyploid (up to octoploid) use (-P|Polyploid)

Usage:

	-h|--help 
	-D|Diploid Species genome.fa 
	-D|Diploid Species genome.fa ChrString 
	-P|Polyploid -abbr Species -g genome.fa -chr ChrString -p1 Ghir_A -p2 Ghir_D 
	-P|Polyploid -abbr Species -g genome.fa -p1 Ghir_A -p2 Ghir_D 
	-V | --version

Example:

	LTRfind -D A.thaliana /path/A.thaliana.fa 
	LTRfind Diploid A.thaliana /path/A.thaliana.fa chr
	LTRfind -P -abbr Species -g genome.fa -chr ChrString -p1 Ghir_A -p2 Ghir_D 
	LTRfind -P -abbr Species -g genome.fa -chr ChrString -p1 Ghir_A -p2 Ghir_D -p3 Ghir_C 
	LTRfind -P -abbr Species -g genome.fa -chr ChrString -p1 Ghir_A -p2 Ghir_D -p3 Ghir_C -p4 Ghir_B
	LTRfind Polyploid -abbr Species -g genome.fa -p1 Ghir_A -p2 Ghir_D
	LTRfind -D A.thaliana ${PWD}/example/LTRfind/Ath.Chr4_5.fa chr

Note:

In run -D|Diploid ，if you input the ChrString , will extract the Chromosome sequence for LTR and LAI (not analysis the scaffold sequence). Input genome file must be .fa,not support .fa.gz.

the genome.fa sequence ID should be less then 13 character. if your file is report over 13 character, can use seqkit seq -i genome.fa >new.genome.fa to extract the ID and delete the info after the ID

if your input genome.fa ChrID may be like >1,in this case, you cannot use the ChrStr field to extract chromosome data and filter scaffold. It is recommended to use LTRfind after filtering by yourself. The ID length of genome.fa cannot exceed 13 characters.

Visualization

if you jsut want use the visualization module visual_LAI and visual_LTR.

visual_LAI.R used only for chomosome, so if please delete the rows which is contain contigs and scaffolds in *.out.LAI

Rscript visual_LAI.R genome.out.LAI speciesname
Rscript visual_LAI.R genome.pass.list speciesname

3. Input

Diploid:

• speciesname
• genome.fa
• ChrStr (can be omitted)

Polyploid:

• speciesname
• genome.fa
• ChrStr (If you do not provide this parameter, it is assumed that the genome file you provide only contains the sequence of the chromosome.)
• p1
• p2
• p3
• p4

Example:

Polyploid:

for Ghirsutum.fasta head Ghirsutum.fasta

>Ghir_A01
TAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAA
ACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACC
CTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTA
AACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAAC
CCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAAACCCTAAACCCTAAACCC
TAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAA
ACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACC
CTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTA
AACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAAC
...
>Ghir_D01
AAACCCTAAACCCTAAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAAACCCTAAACCCTAAACCCTAAACCCTA
AACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAAC
CCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCT
AAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAA
CCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCCTAAACCC

should be run like this:

• LTRfind -P -abbr G.hirsutum -g Ghirsutum.fasta -chr Ghir_ -p1 Ghir_A -p2 Ghir_D
• LTRfind -u 7.0e-9 -P -abbr G.hirsutum -g Ghirsutum.fasta -chr Ghir_ -p1 Ghir_A -p2 Ghir_D

Diploid

head Arabidopsis_thaliana.fa

>chr01 dna:chromosome chromosome:TAIR10:1:1:30427671:1 REF
CCCTAAACCCTAAACCCTAAACCCTAAACCTCTGAATCCTTAATCCCTAAATCCCTAAAT
CTTTAAATCCTACATCCATGAATCCCTAAATACCTAATTCCCTAAACCCGAAACCGGTTT
CTCTGGTTGAAAATCATTGTGTATATAATGATAATTTTATCGTTTTTATGTAATTGCTTA
TTGTTGTGTGTAGATTTTTTAAAAATATCATTTGAGGTCAATACAAATCCTATTTCTTGT
GGTTTTCTTTCCTTCACTTAGCTATGGATGGTTTATCTTCATTTGTTATATTGGATACAA

bash LTRfind -D A.thaliana Arabidopsis_thaliana.faor bash LTRfind -D A.thaliana Arabidopsis_thaliana.fa chr

Note

The string of the first parameter species must be a complete prefix of species.fa or species.fasta, otherwise an error will be reported. For example:

LTRfind -D test.222 test222.fa Chr
LTRfind -D Test_genome Test_genome.fasta
LTRfind -t 16 -D Test_genome Test_genome.fasta
LTRfind -u 1.8e9 -D Test_genome Test_genome.fasta
LTRfind -u 2e-9 -t 16 -D Test_genome Test_genome.fasta

4. Output

• All the output is the same with LTR_retriever.
• For Chinese, there are detail info links in Chinese.
• *.out.LAI

Chr     From    To      Intact  Total   raw_LAI LAI
whole_genome    1       2345845174        0.0191  0.1304  14.64   11.51
Chr1     1       3000000 0.0092  0.1263  11.32    10.19

关于输出的LAI值的问题

该文件的第1行，显示的分别是该物种的基因组总长度是2.3458G,完整的LTR-RTs占基因组比例是1.91%，总共的LTR占基因组的比例是13.04%， 如果一个物种总的LTR占比低于5%或完整的LTR占比低于0.1%，则LAI软件计算出的LAI值不可信。

关于版本问题

目前兼容的是LTR_retriever v2.9.6，不兼容LTR_retriever V3.0.1,因为新版本的更新了LTR时间的默认算法，所以会输出很多NA。

5. Author & Version

• Build date:2021.06.17
• Last update: 2025.06.13
• Version: 0.4
• Author: Mol Chai
• Email: chaimol@163.com

6. Update info

(2025/06/13) update to Version 0.4

• Modify the "LTRfind" and "config.ini".The user is required to provide the full address of Rscript
• Added more detection steps to output error information in a timely manner.
• Added a multi-threaded script LTR_HARVEST_parallel, which can be used to call gt to make it run faster.
• Added two command line parameters -u and -t. -u is used to control the miu evolution rate, the default value is 7e-9 in config.ini, -t is the number of threads used, the default value is 36 in config.ini.

(2022/03/04) update to Version 0.3.1

• Add visual module getLTRpeak.R.
• Add example for test.
• The test passed the module for diploid! Fixed some bugs!

(2022/03/02) update to Version 0.3.0

• Add config.ini file for user control the input software path .
• Add the parameter threads to specify the number of cpus and the parameter miu to specify the differentiation speed.
• Add visual module visul_LTR.R and visual_LAI.R

(2021/11/19)

Modify the value of ltrharvest -similar 90 to ltrharvest -similar 85. In order to ensure that LTR_harvest and LTR_FINDER_parallel use the same similarity value (0.85).

(2021/11/05)

Using LTR_FINDER_parallel instead of LTR_FINDER, the running speed will be greatly improved.