GBSX: a toolkit for experimental design and demultiplexing genotyping by sequencing experiments
Genotyping by Sequencing is an emerging technology for cost effective variant discovery and genotyping. However, current analysis tools do not fulfill all experimental design and analysis needs.
GBSX is a package of tools to first aid in experimental design, including choice of enzymes and barcode design. Secondly, it provides a first analysis step to demultiplex samples using in-line barcodes, providing fastq files that can easily be plugged into existing variant analysis pipelines.
The perl script for in silico digests and the compiled program for all other analyses can be found in the releases directory. The latest directory has the latest version. However, previous versions are still available. The complete source code can be found in the src directory. Example data and results for the tool can be found in the example directory.
All parts of this tool is licenced under GPLv3.
A copy of this licence is included under LICENSE.
Center for Human Genetics
UZ – KU Leuven
Herestraat 49 PO box 602
B-3000 Leuven, Belgium
We ask that you cite this paper if you use GBSX in work that leads to publication.
Herten, Koen and Hestand, Matthew S. and Vermeesch, Joris R. and Van Houdt, Jeroen KJ (2015) GBSX: a toolkit for experimental design and demultiplexing genotyping by sequencing experiments BMC Bioinformatics 2015, 16:73 doi:10.1186/s12859-015-0514-3
Genotyping By Sequencing demultipleXing toolkit (GBSX) is a toolkit with an inline barcode demultiplexer for usage in the analysis of single read or paired-end genotyping by sequence (GBS) data, a barcode generator, a barcode discovery tool, and a restriction enzyme predictor. GBSX can easily be incorperated as a preceding analysis step for already deployed SNP pipelines.
Restriction Enzyme Predictor
`-l` read length
`-f` file of reference fasta file location(s)
-eenzyme name to use (default: Enzyme)
`-g` genome name to use in bed file name (default: genome)
`-n` minimum size fragments to include (default: 100)
`-m` maximum size fragments to use (default: 1000)
`-E` second enzyme name to use (default: Enzyme2)
`-D` digest sequence for a second enzyme (default: not declared)
`-R` digest sequence for a third enzyme (default: not declared)
-bthe number of barcodes needed
-ethe enzyme used for the experiment
-efthe enzyme file. This option adds new enzymes.
The file must be tab delimited: First column the enzyme name, second column the cutsites remains (comma separated) (example: enzyme ApeKI and restriction site G^CWGC: "ApeKI \tab CAGC,CTGC").
-nbthe maximum number of bootstraps that must be executed. (optional, standard 10000). By the start of a new bootstrap a complete new design is made. The best scored design (most random barcodes and best scored bases distribution is kept as result)
-btthe number of barcode tries. (standard 20) If a random barcode does not fit into the current design try this number of times with a new random barcode before restarting the bootstrap.
-othe output directory (standard current working directory)
-ustry to find the ultime match: the best barcode combination with the best bases distribution (standard false) true: continue even when the right number of barcodes is found.
-bfa file with all barcodes that are used as basic set (this file is one of the possible output files)
-nfa file with all barcodes that may not be used in the design. If this file contains barcodes that are also found in the basic set file, these barcodes will be replaced in the design by new random barcodes.
This program demultiplexes fastq or fastq.gz files obtained from sequencing with
Like used in GBS, RAD, ... protocols.
These parameters are mandatory:
-f1the name and path of the fastq or fastq.gz file to demultiplex
-ithe name and path of the info file. This is a tab delimeted file without headings, with three (or more) columns: sample, sequence of the barcode, name of the enzyme, name of the second enzyme (optional, can be an empty string), the second barcode (optional, can be an empty string),mismatches for the barcode (optional)
These parameters are optional:
-f2the name of the second fastq or fastq.gz file (only with paired-end sequencing)
-othe name of the output directory (standard the directory of the call)
-tthe number of threads to use (standard 1)
-lfuse long file names (standard false) filename is standard the sample name, long file names is sample name _ barcode _ enzyme
-radif the data is rad data or not (-rad true for RAD data, -rad false for GBS data) standard false (GBS)
-gzipthe input and output are/must be gziped (.gz) (standard false: input and output are .fastq, if true this is .fastq.gz)
-mbthe allowed mismatches in the barcodes (overrides the option -m)
-methe allowed mismatches in the enzymes (overrides the option -m)
-minslthe minimum allowed length for the sequences (standard 0, rejected sequences are found in the stats for each sample in the rejected.count column. The sequences are found untrimmed in the undetermined file.)
-nkeep sequences where N occurs as a "nucleotide" (standard true)
-cathe common adaptor used in the sequencing (standard (only first piece) AGATCGGAAGAGCG) currently only used for adaptor ligase see -al and when -rad is true) (minimum length is 10)
-sthe posible distance of the start. This is the distance count from the start of the read to the first basepair of the barcode or enzyme (standard 0, maximum 20)
-kcKeep the enzyme cut-site remains (standard true) (example: enzyme ApeKI and restriction site G^CWGC: "ApeKI \tab CAGC,CTGC")
-eaAdd enzymes from the given file (keeps the standard enzymes, and add the new) (enzyme file: no header, enzyme name tab cutsites (multiple cutsites are comma separeted)) (only use once, not use -er) (example: enzyme ApeKI and restriction site G^CWGC: "ApeKI \tab CAGC,CTGC")
-erReplace enzymes from the given file (do not keep the standard enzymes) (enzyme file: no header, enzyme name tab cutsites (multiple cutsites are comma separeted)) (only use once, not use -ea)
-scbUse self correcting barcodes (barcodes created by the barcodeGenerator) (standard false)
-malgthe used algorithm to find mismatches and indels, possible algorithms:
hammings (Standard)Checks for mismatches (no indels)
knuthFaster than hammings, but can miss some locations
indelmisChecks for mismatches and indels, the barcode/enzyme/ adaptor with the least errors (mismatches or indels) is taken
misindelChecks for mismatches and indels, the mismatches are supperior to the indels (faster than indelmis, but errors can be higher)
-qthe kind of quality scores used in the fastq file (including how phred scores are encoded):
Possible Standard Enzymes for the info file: (NA is no enzyme)
This program searches for possible barcodes and barcode enzyme combinations.
Designed for the discovery of sequencing errors, or unused barcodes when a large proportion of the demultiplex is undetermined.
-f1the name of the input file (mandatory)
-minthe minimum length of the barcode (standard 6)
-maxthe maximum length of the barcode (standard 10)
-gzipuse gzip files as input and output (standard false)
-othe output directory (standard the directory of execution)
-eaAdd enzymes from the given file (keeps the standard enzymes, and add the new) (enzyme file: no header, enzyme name tab cutsites (multiple cutsites are comma separeted)) (example: enzyme ApeKI and restriction site G^CWGC: "ApeKI \tab CAGC,CTGC") (only use once, not use -er)
-erReplace enzymes from the given file (do not keep the standard enzymes) (enzyme file: no header, enzyme name tab cutsites (multiple cutsites are comma separeted)) (example: enzyme ApeKI and restriction site G^CWGC: "ApeKI \tab CAGC,CTGC") (only use once, not use -ea)
-barminThe minimum occurance of a barcode before it is shown in the results (standard: 200)
-barmaxThe maximum occurance of barcodes shown in the output (increasing this number will increase ram usage, but gives a slightly better result) (standard: 100)
-barmisThe percentage of mismatches that may occure between barcodes (integer between 1 and 10) (standard: 10)
See the Tutorial file and the example folder.
- The original version
- While demultiplexing, the number of demultiplexed reads are shown for every 100000 reads
- Possible to simulate and demultiplex dual barcode experiments (in paired end modus only)
- Updated barcode recognition for paired end modus in the demultiplexer: when a read can be assigned to multiple samples, the read is considered as unvalid (previous was first sample)
- Updated output and stats for dual barcode experiments
- Updated barcode recognition for single read modus in the demultiplexer: when a read can be assigned to multiple samples, the read is considered as unvalid (previous was first sample)
- On request added the enzyme AvaII
- Update adaptor ligase finding algorithm
- Removed unneeded, confusing parameters -cc and -al
- Removed unused code
- Update digest (removed possible input file parsing error)
- Updated Single Read Demultiplexing
- Deleted Demultiplexer option -m
- Code Clean-up
- Removal of the filelock class, ReentrantLock is now used
- The demultiplexing is now multithreaded: use -t NumberOfThreads