nanotei

A pipeline for detection of non-reference transposon insertions using Nanopore data

Installation

nanotei depends on several programs and python modules:

  - python =3.7.5
  - pysam =0.17.0
  - samtools =1.14
  - minimap2 =2.22
  - biopython =1.76
  - statistics =1.0.3.5
  - pandas =0.24.2

The best way to install nanotei is to use nanotei.yaml file to create conda environment

 conda env create --file nanotei.yaml

How to run

nanotei is written in python and can be run from command line.

  usage: nanotei.py [-h] [-q MAP_Q] [-mlc MIN_LEN_CLIPPED]
                  [-md MIN_MERGE_DISTANCE] [-samtp SAMTOOLS_PATH]
                  [-mm2 MINIMAP2_PATH] [-bt BEDTOOLS]
                  [-ovt OVERLAP_WITH_ORIGTE] [-mpv MINPVALUE] [--fpv]
                  [-mrs MIN_READ_SUPPORT] [--bed]
                  bam_file fastq genome_fasta outfolder bed_TE outtab

Script to find regions of insertions using Nanopore reads

positional arguments:

bam_file path to the sorted bam file with mapped reads to the genome

fastq path to fastq file of reads

genome_fasta path to fasta file of genome reference

outfolder path to the output folder

bed_TE path to the bed file with annotated TE

outtab name of the output table

optional arguments:

-h, --help show this help message and exit

-q MAP_Q, --map_q MAP_Q minimum mapping quality

-mlc MIN_LEN_CLIPPED, --min_len_clipped MIN_LEN_CLIPPED minimum length of the clipped part

-md MIN_MERGE_DISTANCE, --min_merge_distance MIN_MERGE_DISTANCE minimum distance between two reference positions of clipping to be merged into one

-samtp SAMTOOLS_PATH, --samtools_path SAMTOOLS_PATH path to samtools program

-mm2 MINIMAP2_PATH, --minimap2_path MINIMAP2_PATH path to minimap2 program

-bt BEDTOOLS, --bedtools BEDTOOLS path to minimap2 program

-ovt OVERLAP_WITH_ORIGTE, --overlap_with_origTE OVERLAP_WITH_ORIGTE minimum portion of origin TE overlapped with clipped repa alignemnt

-mpv MINPVALUE, --minpvalue MINPVALUE minimum pvalue to filter (it is used with --fpv). pvalue here reflects the probability that the TEI is OK (so usually it should be > 0.05) --fpv filter by pvalue. Note! If it is NOT set then TEIs with low (than expected by Poisson distribution) read number will be reported (including somatic TEIs). This option may significantly reduce calculation time.

-mrs MIN_READ_SUPPORT, --min_read_support MIN_READ_SUPPORT minimum read support for filtering

--bed output bed file with TEIs

input files

bam file can be obtained by read mapping to the reference genome by minimap2 (minimap2 -ax map-ont -t 150 genome.fasta reads.fastq > file.sam) followed by sam to bam conversion (samtools view -Sb file.sam > file.bam) and sorting (samtools sort -o sorted_file.bam file.bam)

bed_TE nanotei requires 4-column bed file without header: chromosome_id, start, end, TEid.

Example of how to run nanotei

get copy of githup repository and change the directory to test

git clone https://github.com/Kirovez/nanotei.git
cd ./nanotei/test

1. download TAIR10 genome from NCBI and move it to test directory (it must be current directory)

TAIR10.1 repository: https://www.ncbi.nlm.nih.gov/assembly/GCF_000001735.4#/def_asm_Primary_Assembly

unpack tar and unzip .gz in the folder. Then move GCF_000001735.4_TAIR10.1_genomic.fna file to test directory (current directory). Examples of the commands:

tar -xvf genome_assemblies_genome_fasta.tar

gunzip ./ncbi-genomes-2021-11-23/GCF_000001735.4_TAIR10.1_genomic.fna.gz

mv ./ncbi-genomes-2021-11-23/GCF_000001735.4_TAIR10.1_genomic.fna .

2. download Nanopore reads of ddm1 plant

wget ftp://ftp.sra.ebi.ac.uk/vol1/fastq/SRR147/094/SRR14765194/SRR14765194_1.fastq.gz 
## gunzip
gunzip SRR14765194_1.fastq.gz

3. TE annotation file that was build by Panda and Slotkin, 2020 must be already in the test directory. The chromosome ids were changed according to the NCBI chromosome ids. The file is available on GitHub:

wget https://raw.githubusercontent.com/Kirovez/nanotei/master/test/TE_31189_Slotkin_NCBI_id.bed

4. map the reads to the genome using minimap2, then convert sam file to bam, sort and index:

mapping

minimap2 -ax map-ont -t 100 GCF_000001735.4_TAIR10.1_genomic.fna SRR14765194_1.fastq > SRR14765194_vs_TAIR10.sam

convert sam to bam

samtools view -Sb SRR14765194_vs_TAIR10.sam > SRR14765194_vs_TAIR10.bam 

sort bam file

samtools sort -o sorted_SRR14765194_vs_TAIR10.bam -@ 100 SRR14765194_vs_TAIR10.bam 

index

samtools index -@ 100 sorted_SRR14765194_vs_TAIR10.bam

5. run nanotei. The resulted files will appear in the current directory (it will be changed in later versions).

python3 ../nanotei.py ./sorted_SRR14765194_vs_TAIR10.bam ./SRR14765194_1.fastq GCF_000001735.4_TAIR10.1_genomic.fna ./out_nanotei ./TE_31189_Slotkin_NCBI_id.bed SRR14765194.nanotei --fpv --bed --minpvalue 0.05 -ovt 0.3

an example ot output table and bed file can be found in out_example folder. Note: TEIs in output folder for ddm1 includes ddm1 specific TEIs and TEIs that resulted from assembly artefacts (~44 TEIs, see nanotei paper for details)

Output table

The output nanotei table contains the following columns:

TEIcoord - the coordinates of TEI: chromosome:startTEI..endTEI

supp.reads - the number of reads supporting this TEI (reads with clipped ends + reads with insertions)

pvalue - the probability that this insertion is real based on the number of reads supporting TEI and the genome coverage distribution (so 1 is absolutely OK and 0 - is bad TEI).

isTEIregioIsOK - it is FALSE if the number of reads in the region flanking TEI is too high based on expectation from genome coverage distribution

TE_reads - it shows the number of mapped sequences of clipped parts and insertion parts of the reads overlaped with TEs from bed file.

chosenTE - a TE with the highest number of mapped clipped parts and insertion parts of the reads

TE.chr - chromosome name of chosen TE

TE.start - chromosome start of chosen TE

TE.end - chromosome end of chosen TE

nanotei citation

If you use nanotei, please, cite our paper: https://doi.org/10.3390/plants10122681