Mimeo

Table of contents

• Modules
• Installing Mimeo
• Example usage
• Standard options
  • mimeo-self
  • mimeo-x
  • mimeo-map
  • mimeo-filter
• Alternative alignment engines
• License

Modules

Mimeo comprises three tools for parsing repeats from whole-genome alignments:

mimeo-self

Internal repeat finder. Mimeo-self aligns a genome to itself and extracts high-identity segments above a coverage threshold. This method is less sensitive to disruption by indels and repeat-directed point mutations than kmer-based methods such as RepeatScout. Reported annotations indicate overlapping segments above the coverage threshold, mimeo-self does not attempt to separate nested repeats. Use this tool to identify candidate repeat regions for curated annotation.

mimeo-x

Cross-species repeat finder. A newly acquired or low-copy transposon may slip past copy-number based annotation tools. Mimeo-x searches for features which are abundant in an external reference genome, allowing for annotation of complete elements as they occur in a horizontal-transfer donor species, or of conserved coding segments of related transposon families.

mimeo-map

Find all high-identity segments shared between genomes. Mimeo-map identifies candidate horizontally transferred segments between sufficiently diverged species. When comparing isolates of a single species, aligned segments correspond to directly homologous sequences and internally repetitive features.

Intra/Inter-genomic alignments from Mimeo-self or Mimeo-x can be reprocessed with Mimeo-map to generate annotations of unfiltered/uncollapsed alignments. These raw alignment annotations can be used to interrogate repetitive-segments for coverage breakpoints corresponding to nested transposons with differing abundances across the genome.

mimeo-filter

An additional tool mimeo-filter is now included to allow post-filtering of SSR-rich sequences from FASTA formatted candidate-repeat libraries.

Installing Mimeo

Requirements:

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

Install from PyPi:

pip install mimeo

Clone and install from this repository:

git clone https://github.com/Adamtaranto/mimeo.git && cd mimeo && pip install -e .

Example usage

mimeo-self

Annotate features in genome A which are > 100bp and occur with >= 80% identity at least 3 times on other scaffolds OR at least 4 times on the same scaffold.

mimeo-self --adir data/A_genome_Split --afasta data/A_genome.fasta \
-d MS_outdir --gffout A_genome_Inter3_Intra4_id80_len_100.gff3 \
--outfile A_genome_Self_Align.tab --label A_Rep3 --prefix A_Self --minIdt 80 \
--minLen 100 --minCov 3 --intraCov 4 --strictSelf

Output:

• MS_outdir/A_genome_Inter3_Intra4_id80_len_100.gff3
• MS_outdir/A_genome_Self_Align.tab
• data/A_genome_Split/*.fa

mimeo-x

Annotate features in genome A which are > 100bp and occur with >= 80% identity at least 5 times in genome B.

mimeo-x --afasta data/A_genome.fasta --bfasta data/B_genome.fasta \
-d MX_outdir --gffout B_Rep5_in_A.gff3 --outfile B_Reps_in_A_id80_len100.tab \
--label B_Rep5 --prefix B_Rep5 --minIdt 80 --minLen 100 --minCov 5

Output:

• MX_outdir/B_Rep5_in_A.gff3
• MX_outdir/B_Reps_in_A_id80_len100.tab

mimeo-map

Annotate features in genome A which are > 100bp and occur with >= 90% identity in genome B. No coverage filter, all alignments are reported.

mimeo-map --afasta data/A_genome.fasta --bfasta data/B_genome.fasta \
-d MM_outdir --gffout B_in_A_id90.gff3 --outfile B_in_A_id90.tab \
--label B_90 --prefix B_90 --minIdt 90 --minLen 100 

Output:

• MM_outdir/B_in_A_id90.gff3
• MM_outdir/B_in_A_id90.tab

mimeo-map + SSR filter

Annotate features in genome A which are > 100bp and occur with >= 98% identity in genome B. Reuse B to A-genome alignment from the previous run.

Filter out hits which are >= 40% tandem repeats. Write filtered hits as tab file and GFF3 annotation.

mimeo-map --afasta data/A_genome.fasta --bfasta data/B_genome.fasta \
-d MM_outdir --gffout B_in_A_id98_maxSSR40.gff3 --outfile B_in_A_id98.tab \
--label B_98 --prefix B_98 --minIdt 98 --minLen 100 \
--recycle --maxtandem 40 --writeTRF

Output:

• MM_outdir/B_in_A_id98_maxSSR40.gff3
• MM_outdir/B_in_A_id98.tab.trf

mimeo-filter

Filter sequences comprised of >= 40% short tandem repeats from a multifasta library of candidate transposons.

mimeo-filter --infile data/candidate_TEs.fa

Output:

• candidate_TEs_filtered.fa

Standard options

mimeo-self

Usage: mimeo-self [-h] [--adir ADIR] [--afasta AFASTA] [-r] [-d OUTDIR]
                  [--gffout GFFOUT] [--outfile OUTFILE] [--verbose]
                  [--label LABEL] [--prefix PREFIX] [--lzpath LZPATH]
                  [--bedtools BEDTOOLS] [--minIdt MINIDT] [--minLen MINLEN]
                  [--minCov MINCOV] [--hspthresh HSPTHRESH]
                  [--intraCov INTRACOV] [--strictSelf]

Internal repeat finder. Mimeo-self aligns a genome to itself and extracts
high-identity segments above a coverage threshold.

Optional arguments:
  -h, --help      Show this help message and exit.
  --adir          Name of the directory containing sequences from the genome.
                  Write split files here if providing genome as
                  multifasta.
  --afasta        Genome as multifasta.
  -r, --recycle   Use existing alignment "--outfile" if found.
  -d , --outdir   Write output files to this directory. (Default: cwd)
  --gffout        Name of GFF3 annotation file.
  --outfile       Name of alignment result file.
  --verbose       If set report LASTZ progress.
  --label         Set annotation TYPE field in gff.
  --prefix        ID prefix for internal repeats.
  --lzpath        Custom path to LASTZ executable if not in $PATH.
  --bedtools      Custom path to bedtools executable if not in $PATH.
  --minIdt        Minimum alignment identity to report.
  --minLen        Minimum alignment length to report.
  --minCov        Minimum depth of aligned segments to report repeat
                  feature.
  --hspthresh     Set HSP min score threshold for LASTZ.
  --intraCov      Minimum depth of aligned segments from the same scaffold
                  to report feature. Used if "--strictSelf" mode is
                  selected.
  --strictSelf    If set process same-scaffold alignments separately
                  with the option to use higher "--intraCov" threshold.
                  Sometimes useful to avoid false repeat calls from
                  staggered alignments over SSRs or short tandem
                  duplication.

mimeo-x

Usage: mimeo-x [-h] [--adir ADIR] [--bdir BDIR] [--afasta AFASTA]
               [--bfasta BFASTA] [-r] [-d OUTDIR] [--gffout GFFOUT]
               [--outfile OUTFILE] [--verbose] [--label LABEL]
               [--prefix PREFIX] [--lzpath LZPATH] [--bedtools BEDTOOLS]
               [--minIdt MINIDT] [--minLen MINLEN] [--minCov MINCOV]
               [--hspthresh HSPTHRESH]

Cross-species repeat finder. Mimeo-x searches for features which are abundant
in an external reference genome.

Optional arguments:
  -h, --help      Show this help message and exit.
  --adir          Name of the directory containing sequences from A genome.
  --bdir          Name of the directory containing sequences from B genome.
  --afasta        A genome as multifasta.
  --bfasta        B genome as multifasta.
  -r, --recycle   Use existing alignment "--outfile" if found.
  -d , --outdir   Write output files to this directory. (Default: cwd)
  --gffout        Name of GFF3 annotation file.
  --outfile       Name of alignment result file.
  --verbose       If set report LASTZ progress.
  --label         Set annotation TYPE field in GFF.
  --prefix        ID prefix for B-genome repeats annotated in A-genome.
  --lzpath        Custom path to LASTZ executable if not in $PATH.
  --bedtools      Custom path to bedtools executable if not in $PATH.
  --minIdt        Minimum alignment identity to report.
  --minLen        Minimum alignment length to report.
  --minCov        Minimum depth of B-genome hits to report feature in
                  A-genome.
  --hspthresh     Set HSP min score threshold for LASTZ.

mimeo-map

Usage: mimeo-map [-h] [--adir ADIR] [--bdir BDIR] [--afasta AFASTA]
                 [--bfasta BFASTA] [-r] [-d OUTDIR] [--gffout GFFOUT]
                 [--outfile OUTFILE] [--verbose] [--label LABEL]
                 [--prefix PREFIX] [--keeptemp] [--lzpath LZPATH]
                 [--minIdt MINIDT] [--minLen MINLEN] [--hspthresh HSPTHRESH]
                 [--TRFpath TRFPATH] [--tmatch TMATCH] [--tmismatch TMISMATCH]
                 [--tdelta TDELTA] [--tPM TPM] [--tPI TPI]
                 [--tminscore TMINSCORE] [--tmaxperiod TMAXPERIOD]
                 [--maxtandem MAXTANDEM] [--writeTRF]

Find all high-identity segments shared between genomes.

Optional arguments:
  -h, --help      Show this help message and exit.
  --adir          Name of the directory containing sequences from A genome.
  --bdir          Name of the directory containing sequences from B genome.
  --afasta        A genome as multifasta.
  --bfasta        B genome as multifasta.
  -r, --recycle   Use existing alignment "--outfile" if found.
  -d, --outdir    Write output files to this directory. (Default: cwd)
  --gffout        Name of GFF3 annotation file. If not set, suppress
                  output.
  --outfile       Name of alignment result file.
  --verbose       If set report LASTZ progress.
  --label         Set annotation TYPE field in GFF.
  --prefix        ID prefix for B-genome hits annotated in A-genome.
  --keeptemp      If set does not remove temp files.
  --lzpath        Custom path to LASTZ executable if not in $PATH.
  --minIdt        Minimum alignment identity to report.
  --minLen        Minimum alignment length to report.
  --hspthresh     Set HSP min score threshold for LASTZ.
  --TRFpath       Custom path to TRF executable if not in $PATH.
  --tmatch        TRF matching weight.
  --tmismatch     TRF mismatching penalty.
  --tdelta        TRF indel penalty.
  --tPM           TRF match probability.
  --tPI           TRF indel probability.
  --tminscore     TRF minimum alignment score to report.
  --tmaxperiod    TRF maximum period size to report.
  --maxtandem     Max percentage of an A-genome alignment which may be masked by TRF. 
                  If exceeded, the alignment will be discarded.
  --writeTRF      If set write TRF filtered alignment file for use with
                  other mimeo modules.

mimeo-filter

Usage: mimeo-filter [-h] --infile INFILE [-d OUTDIR] [--outfile OUTFILE]
                    [--keeptemp] [--verbose] [--TRFpath TRFPATH]
                    [--tmatch TMATCH] [--tmismatch TMISMATCH]
                    [--tdelta TDELTA] [--tPM TPM] [--tPI TPI]
                    [--tminscore TMINSCORE] [--tmaxperiod TMAXPERIOD]
                    [--maxtandem MAXTANDEM]

Filter SSR containing sequences from FASTA library of repeats.

Optional arguments:
  -h, --help            Show this help message and exit.
  --infile              Name of the directory containing sequences from A genome.
  -d, --outdir          Write output files to this directory. (Default: cwd)
  --outfile             Name of alignment result file.
  --keeptemp            If set does not remove temp files.
  --verbose             If set report LASTZ progress.
  --TRFpath             Custom path to TRF executable if not in $PATH.
  --tmatch              TRF matching weight
  --tmismatch           TRF mismatching penalty.
  --tdelta              TRF indel penalty.
  --tPM                 TRF match probability.
  --tPI                 TRF indel probability.
  --tminscore           TRF minimum alignment score to report.
  --tmaxperiod          TRF maximum period size to report. Note: Setting this
                        score too high may exclude some LTR retrotransposons.
                        Optimal len to exclude only SSRs is 10-50bp.
  --maxtandem           Max percentage of a sequence which may be masked by
                        TRF. If exceeded, the element will be discarded.

Importing alignments

Whole genome alignments generated by alternative tools (i.e. BLAT) can be provided to any of the Mimeo modules as a tab-delimited file with the columns:

[1]   name1     = Name of target sequence in genome A
[2]   strand1   = Strand of alignment in target sequence
[3]   start1    = 5-prime position of alignment in target (lower value irrespective of strand)
[4]   end1      = 3-prime position of alignment in target (higher value irrespective of strand)
[5]   name2     = Name of source sequence in genome B
[6]   strand2   = Strand of alignment in source
[7]   start2+   = 5-prime position of alignment in source (lower value irrespective of strand)
[8]   end2+     = 3-prime position of alignment in source (higher value irrespective of strand)
[9]   score     = Alignment score as int
[10]  identity  = Identity of alignment as float

File should be sorted by columns 1,3,4

License

Software provided under MIT license.