Tinscan: TE-Insertion-Scanner

Scan whole genome alignments for transposon insertion signatures.

Table of contents

• Algorithm overview
• Options and usage
  • Installing Tinscan
  • Example usage

Algorithm overview

1. Perform gapped and chained whole genome alignment of query genome B onto target genome A.
2. Where two aligned segments are contiguous in B (or separated by no more than --qGap), and
3. Separated by an insertion in the range --minInsert:--maxInsert in A, and
4. At least one flanking alignment in A satisfies the threshold --minIdent, and differs from its mate by no more than --maxIdentDiff %
5. Log flanks and candidate insertion.
6. Attempt to infer TSDs from the internal overlap of flanking alignments in B genome.

Options and usage

Installing Tinscan

Requirements:

• LASTZ genome alignment tool from the Miller Lab, Penn State.
• Biopython

You can set up a conda environment with the required dependencies using the YAML files in this repo:

For ARM64 (Apple Silicon Macs) create a virtual intel env.

# For ARM64 Macs only
conda env create -f env_osx64.yml
conda activate tinscan-osx64

For all other operating systems use environment.yml

conda env create -f environment.yml
conda activate tinscan

With the conda env active you can now install tinscan.

1. Install from PyPi.

pip install tinscan

2. For the latest development version, clone and install from this repository.

git clone https://github.com/Adamtaranto/TE-insertion-scanner.git && cd TE-insertion-scanner && pip install -e ".[tests]"

Example usage

Find insertion events in genome A (target) relative to genome B (query).

Prepare Input Genomes

Split A and B genomes into two directories containing one scaffold per file. Check that sequence names are unique within genomes.

tinscan-prep --adir data/A_target_split --bdir data/B_query_split\
-A data/A_target_genome.fasta -B data/B_query_genome.fasta 

Output: data/A_target_split/.fa data/B_query_split/.fa

Align Genomes

Align each scaffold from genome B onto each genome A scaffold. Report alignments with >= 60% identity and length >= 100bp.

tinscan-align --adir data/A_target_split --bdir data/B_query_split \
--outdir A_Inserts --outfile A_Inserts_vs_B.tab \
--minIdt 60 --minLen 100 --hspthresh 3000

Output: A_Inserts/A_Inserts_vs_B.tab

Note: Alignment tasks can be limited to a specified set of pairwise comparisons where appropriate (i.e. when homologous chromosome pairs are known between assemblies) using the option --pairs.

Comparisons are specified with a tab-delimited text file, where column 1 contains sequence names from genome A, and column 2 contains sequences from genome B.

In the example Chromosome_pairs.txt, Chr A2 has been assembled as two scaffolds (B2, B3) in genome B.

#Chromosome_pairs.txt
A1    B1
A2    B2
A2    B3
A3    B4

Find Insertions

Scan alignments for insertion events and report as GFF annotation of Genome A

tinscan-find --infile A_Inserts/A_Inserts_vs_B.tab \
--outdir A_Inserts --gffOut A_Inserts_vs_B_l100_id80.gff3 \
--maxInsert 50000 --minIdent 80 --maxIdentDiff 20

Output: A_Inserts/A_Inserts_vs_B_l100_id80.gff3

License

Software provided under MIT license.