AmBiVErT
This program is designed for the processing of amplicon based resequencing data generated on second generation sequencing platforms (eg Illumina HiSeq/MiSeq).
AmBiVErT was designed for use in pathology applications but is useful in a variety of amplicon sequencing settings - including CRISPR target site sequencing.
The approach used is to batch reads derived from each amplicon together in a clustering step prior to aligning the sequence to a reference genome.
Steps:
1) Cluster similar reads
2) Align for overlap
3) Align to reference target sequences
4) Consolidate calls
5) Output VCF format callsCreated by Matthew Wakefield and Graham Taylor. Copyright (c) 2013-2017 Matthew Wakefield and The University of Melbourne. All rights reserved.
Installation
AmBiVErT uses a C extension for accelerating the alignment steps. Unfortunately this can complicate the install process.
A wheel distribution is available on the Python Package Index for MacOS 10.11 with Python 3.5 and MacOS 10.12 with Python 3.6. If you are using these versions installation is simply acchieved with:
pip3 install ambivertTo install from source on Mac OS 10.10 - 10.12 with homebrew
brew install --upgrade python3
pip3 install --upgrade pip
pip3 install --upgrade setuptools
pip3 install git+https://github.com/genomematt/AmBiVErT.git
python3 -m ambivert.tests.test_ambivertInstallation from source is known to work on Ubuntu 14.04 & 12.04 and CentOS 5. A Dockerfile is included in the distribution and an image is available (genomematt/ambivert). Some issues may exist with some python virtual environments due to modifications required to avoid a setuptools bug (issue #385).
Our policy is to support use on the latest MacOS, latest LTS Ubuntu and a Docker container.
Usage
$ ambivert --help
usage: ambivert [-h] [-f FORWARD] [-r REVERSE] [-m MANIFEST] [--fasta FASTA]
[--output OUTPUT] [--countfile COUNTFILE]
[--threshold THRESHOLD] [--min_cover MIN_COVER]
[--min_reads MIN_READS] [--min_freq MIN_FREQ]
[--overlap OVERLAP] [--hashtable HASHTABLE]
[--savehashtable SAVEHASHTABLE] [--alignments ALIGNMENTS]
[--prefix PREFIX]AmBiVErT: A program for binned analysis of amplicon data AmBiVErT clusters identical amplicon sequences and thresholds based on read frequency to remove technical errors. Due to sequencing errors occuring with a more random distribution than low frequency variants this approach reduces the number of amplicon products that must be assigned to target regions & assessed for variant calls. AmBiVErT overlaps forward and reverse reads from the same amplicon and preserves local phasing information. Typical running time for first use is several hours, which reduces to less than 10 minutes when the hash table calculated on a previous run is supplied for analysis of subsequent samples with the same amplicons.
optional arguments:
-h, --help show this help message and exit
-f FORWARD, --forward FORWARD
a fastq format file of forward direction amplicon
reads. May be compressed with gzip with the
appropriate suffix (.gz)
-r REVERSE, --reverse REVERSE
a fastq format file of reverse direction amplicon
reads. May be compressed with gzip with the
appropriate suffix (.gz)
-m MANIFEST, --manifest MANIFEST
an Illumina TrueSeq Amplicon manifest file.
--fasta FASTA an fasta amplicon manifest file. Sequences should be
limited to the regions to be called and exclude
primers & adaptors. This file can be provided in
addition to an Illumina manifest to specify additional
off target regions.
--output OUTPUT output of alignments with variants. Default: stdout
--countfile COUNTFILE
output of occurance counts per amplicon. Includes all
counts that map to the reference amplicon. This count
does not include reads that occured at frequencies
below --threshold <default=20>
--threshold THRESHOLD
the minimum occurance threshold. Unique amplicon
sequence variants that occur fewer than threshold
times are ignored. Default 20
--min_cover MIN_COVER
the minimum coverage at a site required to call a
variant. This parameter only has effect if it is >
threshold. Default 0
--min_reads MIN_READS
the minimum number of variant containing reads
required to call a variant. This parameter only has
effect if it is > threshold. Default 0
--min_freq MIN_FREQ the minimum proportion of mutated reads. Default 0.1
(ten percent)
--overlap OVERLAP The minimum overlap required between forward and
reverse sequences to merge. Default 20bp
--hashtable HASHTABLE
Filename for a precomputed hash table of exact matches
of amplicons to references. Generate with
--savehashtable
--savehashtable SAVEHASHTABLE
Output a precomputed hash table that matches amplicons
exactly to references. Used to speed up matching with
--hashtable
--alignments ALIGNMENTS
Print a formatted text version of variant containing
alignments to a file. "-" will print to stderr
--prefix PREFIX Shorthand specification of --forward, --reverse,
--countfile and --outfile. --forward =
<prefix>_R1.fastq.gz --reverse = <prefix>_R2.fastq.gz
--outfile = <prefix>.vcf --countfile = <prefix>.countsAdditional tools for simulating amplicon data
simulate_variants
usage: simulate_variants [-h] [--manifest MANIFEST] [--fasta FASTA]
[--read1 READ1] [--read2 READ2] [--skip_softmasked]
[--position_excludes_softmasked]
[--deletions DELETIONS] [--check_sam CHECK_SAM]A script for simulating paired end reads with variants from an amplicon target file
optional arguments:
-h, --help show this help message and exit
--manifest MANIFEST an Illumina TrueSeq Amplicon manifest file. Default:
None
--fasta FASTA a fasta file of amplicons all in the plus strand
orientation with description lines ">name chromosome
start end" Default: None
--read1 READ1 a fastq output file of forward reads. Default: stdout
--read2 READ2 a fastq output file of reverse reads. Default: stdout
--skip_softmasked dont generate variants in softmasked sequence.
Default: True
--position_excludes_softmasked
Exclude softmasked sequence when calculating start
site of read, cigar and variant detection strings.
Default: True
--deletions DELETIONS
The size deletions to insert instead of point
variants. Default: None
--check_sam CHECK_SAM
a sam file for parsing to identify entries where name
does not match the sam file mapping location and
variant stringstruseq_manifest
usage: truseq_manifest [-h] [--manifest MANIFEST] [--output OUTPUT] [--probes]
[--adaptors] [--with_probes] [--softmask_probes]
[--all_plus]A script for converting Illumina TruSeq Amplicon manifest files to fasta files Produces either a fasta file of target sequences without primers or a file of primer sequences suitable for use by a trimming program (eg Nesoni clip)
optional arguments:
-h, --help show this help message and exit
--manifest MANIFEST an Illumina TruSeq Amplicon manifest file. Default:
stdin
--output OUTPUT a multi fasta output file of sequence targets. Default:
stdout
--probes output only the ULSO and DLSO primer sequences
--adaptors append Illumina adaptor sequences to the primer
sequences
--with_probes append the ULSO and DLSO sequences to the fasta target
sequences
--softmask_probes append the ULSO and DLSO sequences to the fasta target
sequences
--all_plus reorient target sequences so they are all presented on
the plus strand Citing AmBiVErT
Currently AmBiVErT is unpublished, but this repository does have a DOI identifier for each release you can use to cite the code. The DOI for the current release is http://dx.doi.org/10.5281/zenodo.32885