MGEcatcher is an automated pipeline to identify non-fixed mobile genetic elements (MGE) in a set of long read DNA data. MGEcatcher can be used to remove reads containing those MGEs in order to improve genome assembly, or even to survey transposition events in the giving long read DNA sequencing library. MGEcatcher only needs to be provided with a fasta file containing the MGE sequences and the long reads fastq file.

Installation

MGEcatcher is a Bash script, so it doesn't have to be installed. However, it requires the following:

• R (>=4.0.5)
  • GenomicRanges (Bioconductor)
  • IRanges (Bioconductor)
  • ggbio (Bioconductor)
  • dplyr
• Python3 (>=3.8.3)
  • PyFaidx
• Bedtools (>=2.30.0)
• Blastn (>=2.9.0)
• BBMAP (>=38.96)

If you have Conda in your machine, you can easily install all these tools using the env.yml file provided here.

# clone this repo
git clone https://github.com/vhfsantos/MGEcatcher

# enter the directory
cd MGEcatcher

# create the MGEcatcher environment with all dependencies
conda env create -n MGEcatcher -f misc/env.yml

# activate the MGEcatcher env
conda activate MGEcatcher

After, you might want to execute the MGEcatcher script to see if the installation was done correctly:

./MGEcatcher -h

Input Data

MGEcatcher works with minION or PacBio long reads data in fastq format. Besides the fastq file, you will also need to input the fasta file containing the sequences of the MGEs.

If you don't have this MGE seed fasta file, you can try the following:

1. Download the gff annotation file for your reference genome
2. Select only the MGEs features from the raw gff (let's call this output mges-only.gff)
3. Use mges-only.gff as input for the BEDtools' GetFasta program to extract the sequence of these features.
4. (Optional) You may want to cluster the MGE sequences with tools such as CD-HIT

Now you can run MGEcatcher properly, passing the created file as the --mge_seed parameter.

Usage

You must run MGEcatcher for a single fastq file of your library at a time. So you may want to use the --prefix parameter to inform the barcode identifier for the output files.

MGEcatcher --reads <reads.fq> --mge_seed <mge.fa> [--threads <N>] \\
           --output <path/to/output/> --prefix <BC01>

     --reads        Fastq file for basecalled reads to be used as subject
     --mge_seed     Fasta file containing MGE sequences to be used as query
     --output       Output dir name; will be created if does not exist
     --prefix       Prefix for output files (ex.: BC01)
     --threads      Number of threads

Output files

MGEcatcher outputs several files. Inside the _classification folder, you will find the information about the reads containing fixed and unfixed MGEs. Files named -readnames.txt contains only the names of the reads for each classification. Filed named only .txt contains more detailed information about the analysis.

For the .unfixedMGEs.txt file, the column queryname contains the read in which the unfixed MGE was found. The seqnames column contains the names of the reads containing the without-MGE version of the region sequenced by the query. For the fixedMGEs.txt, queryname has the name of the reads whose MGE was also found in the read in the seqnames column.

The .txt files may be used as input for a transposition analysis of your libraries (e.g for comparing the change in the transposition events among the sequenced libraries). Also, the .unfixedMGEs.txt file can be used to remove those reads from the raw fastq file (e.g for improving the genome assembly of the sequenced libraries).

Acknowledgements

This work was founded by Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP - Brazil), grant 2018/25329-9.