find_te_ins
find_te_ins is designed to find Transposon Element (TE) insertions using long reads (nanopore), by alignment directly. (minimap2)
Install
$ git clone https://github.com/bakerwm/find_te_ins.git
$ cd find_te_ins
Change the following variables upon your condition: genome_fa and te_fa in line-10 and line-11;
$ bash run_pipe.sh
run_pipe.sh
Prerequisite
- minimap2 - 2.17-r974-dirty, align long reads to reference genome
- featureCounts - v2.0.0, quantification
- samtools - v1.12, working with BAM files
- python 3.8+
- pysam 0.16.0.1, python module, working with BAM files
Getting Started
1 Prepare input files
- genome_fa - reference genome in fasta format, in script
run_pipe.sh, line-10 - te_fa - TE consensus sequence in fasta format, in script
run_pipe.sh, line-11 - long reads - Long reads from
NanoPoreorPacbio, in fasta or fastq format
2 Run pipe
$ cd ~/work/te_ins
# specify the path of long reads data:
/
$ git clone https://github.com/bakerwm/find_te_ins.git
$ bash find_te_ins/run_pipe.sh <path-to-long-reads>/ results
[1/9] align to reference genome
[2/9] extract raw insertions from BAM, by CIGAR
[3/9] convert raw insertions to fasta format
[4/9] align raw_insertion to transposon
[5/9] extract transposon name for insertions
[6/9] merge raw_insertions by window=100
[7/9] count reads for each insertion
[8/9] save final insertions to file
[9/9] Done!
3 Output
The following files listed below are the output of the pipeline, the TE insertions saved in file *.te_ins.final.bed
$ tree -L 2 results/ONT_sample-1
.
├── ONT_sample-1
│ ├── ONT_sample-1.bam
│ ├── ONT_sample-1.bam.bai
│ ├── ONT_sample-1.raw_ins.bed
│ ├── ONT_sample-1.raw_ins.fa
│ ├── ONT_sample-1.raw_ins.fa.bam
│ ├── ONT_sample-1.raw_ins.fa.bam.bai
│ ├── ONT_sample-1.te_ins.bed
│ ├── ONT_sample-1.te_ins.final.bed
│ ├── ONT_sample-1.te_ins.final.bed6
│ ├── ONT_sample-1.te_ins.gtf
│ ├── ONT_sample-1.te_ins.quant.stderr
│ ├── ONT_sample-1.te_ins.quant.stdout
│ ├── ONT_sample-1.te_ins.quant.txt
│ ├── ONT_sample-1.te_ins.quant.txt.summary
│ ├── ONT_sample-1.te_ins.raw.txt
│ ├── run_minimap2.dm6.stderr
│ └── run_minimap2.dm6_transposon.stderr
...
{sample_name}.te_ins.final.bed
column 1. chr name of reference
column 2. start pos of Insertion
column 3. end pos of Insertion
column 4. insertion name
column 5. a fixed integer [255]
column 6. strand # in current version, not consider the dirction of TE insertions !!!
column 7. name of TE consensus
column 8. length of TE consensus
column 9. proportion of the TE consensus identified
column 10. number of supported reads for the insertion
column 11. number of all reads cover the insertion
column 12. proportion TE supported reads
column 13. type of the TE insertions [full, p3, p5]
{sample_name}.te_ins.raw.txt
column 16 (last column), is the type of TE insertions: [full, p3, p5]
- full, more then cutoff [60%] of the TE consensus were detected
- p3, only the 3' end of the TE consensus were detected
- p5, only the 5' end of the TE consensus were detected
In the .final.bed file, ONLY full TE insertions were saved for further analysis
Change criteria
TE types were defined in run_pipe.sh by anno_te.py, the criteria -c 0.6 could be changed to [0-1] float number based on your condition. see line-100 in file run_pipe.sh
# line-100 of run_pipe.sh
[[ ! -f ${te_ins_txt} ]] && python ${src_dir}/anno_te.py -x ${te_fa_fai} ${te_bam} | sort -k4,4 -k5,5n > ${te_ins_txt}
# change criteria to 0.7
[[ ! -f ${te_ins_txt} ]] && python ${src_dir}/anno_te.py -x ${te_fa_fai} -c 0.7 ${te_bam} | sort -k4,4 -k5,5n > ${te_ins_txt}
# remove te_ins files, and run the command again
$ rm results/ONT_sample-1.te_ins*
$ bash find_te_ins/run_pipe.sh
/ results
How it works?
- extract INSERTIONS
1 Dec 06, 2021
10 Aug 04, 2022
1 Dec 02, 2021
3 Jan 02, 2023
25 Dec 30, 2022
3 Feb 21, 2022
6 Oct 02, 2022
52 Jan 07, 2023
1 Feb 23, 2022
63 Dec 27, 2022
3 Oct 24, 2022
106 Dec 27, 2022
23 May 18, 2022
12 Apr 19, 2022
27 Aug 28, 2022
12 May 08, 2022
1 Dec 04, 2021
1 Dec 07, 2021
25 Dec 12, 2022
1 Nov 10, 2021