In DKMS-LSL/dr2s: Dual redundant reference sequencing
DR2S is a tool to create fully-phased reference grade sequences of heterozygous loci. It is especially designed for working with the highly polymorphic HLA and KIR genes. A detailed description of the algorithm and how it works can be found at bioRxiv.
Input:
The usual inputs to DR2S are fastqs of two sequencing experiments, one of short-reads of shotgun-sequencing (e.g. Illumina), the second of long-reads from third-generation sequencing, e.g. ONT Nanopore or PacBio.
Additionally, a generic reference needs to be give, either as an HLA locus, HLA or KIR allele identifier or path to a fasta file.
Config
Input parameters and paths are either bundled in a config file (yaml or json) or can be provided in an interactive R session.
Possible arguments are described in the function documentation of DR2S::createDR2SConf() ?DR2S::createDR2SConf().
A default config would look like this:
datadir: <path to the data>
outdir: <output directory>
pipeline: SR # SR: with short-reads, LR: only long-reads
format: yaml
longreads: # long-read configuration
type: pacbio # type of long-reads, pacbio or nanopore
dir: pacbio_lowQual # directory containing the fastq files. Either subdirectory of <datadir> or absolute path
shortreads: # short-read configuration
type: illumina # type of short-reads. Usually illumina
dir: illumina # directory containing the fastq files. Either subdirectory of <datadir> or absolute path
sampleId: sample1
locus: HLA-A
reference: HLA-A # Either an HLA or KIR allele identifier, an HLA locus or the path to a reference in fasta format
details: # Any field in details will go to the header of the final fasta.
partnerAllele: another
notes: xx
The yaml config can be read using the readDR2SConf() function.
# library(DR2S)
devtools::load_all()
configFile <- system.file("test/dr2s_conf.yaml", package = "DR2S") # read the example config
config <- readDR2SConf(configFile)
The configuration can also be provided via R in an interactive session or script:
library(DR2S)
config <- createDR2SConf(
sample = "sample3",
locus = "HLA-A",
longreads = list(dir = "nanopore_sampled", type = "nanopore", mapper = "minimap"),
shortreads = list(dir = "Illumina_sampled", type = "illumina", mapper = "rsubread"),
datadir = system.file("testData", package = "DR2S"),
outdir = tempdir(),
reference = "HLA-A"
)
Initialization
The DR2S object must be initialized with a valid config using the InitDR2S() command.
## Initialize the DR2S object
dr2s <- InitDR2S(config)
## The output directory can be cleaned if there remains data from previous attempts using the clear() function. This deletes all the data and re-initialize the project.
clear(dr2s)
## Look what it can tell you
print(dr2s)
InitDR2S() creates the output directory, writes the config as a json file, creates the log and copies the reference.
By default, a folder structure will be created for the output directory consisting of the outdir param, the locus and the sampleId. This can be suppressed by setting createOutdir=FALSE .
The DR2S object is an R6 class, which means it behaves like a reference class. It is always modified in-place and an assignment after the initialization is not necessary.
All methods in the DR2S workflow expect a DR2S object as input and return a DR2S object, which means they can be chained using magrittr's pipe operator %>%.
Initial mapping
The mapInit step creates a first sample-specific consensus sequence by a mapping of the short-reads to the generic reference and subsequent consensus calling. This sequence is still heterozygous. The long- and short-reads are mapped to this sample-specific consensus. Options are drawn from the initial config stored in the DR2S object.
## Run mapInit
dr2s <- mapInit(dr2s)
The mapping or coverage plot is created for long- and short-reads and is stored in the output directory.
imgPath <- file.path(dr2s$getOutdir(), "plot.mapInit.png")
knitr::include_graphics(imgPath)
cluster the long-reads
The clustering of reads based on the long-reads is carried out using the partitionLongreds command.
Heterozygous positions are derived from short-reads of the mapIninit step. Long-reads are clustered based on these heterozygous positions.
## cluster the long-reads
dr2s <- partitionLongreads(dr2s)
Different plots are created to visualize the clustering.
imgPath <- file.path(dr2s$getOutdir(), "plot.partition.png")
knitr::include_graphics(imgPath)
The top panel shows a histogram of the haplotype membership coefficient for assigned haplotypes. Different colors indicate different haplotypes.
The middle panel shows the tree of hierarchical clustering. The blue line indicates the cut-height for haplotype inference. Leafs are single long-reads.
The bottom panel shows assignment of reads assigned to each cluster to all other clusters. This is only useful in cases of more than two haplotypes.
imgPath <- file.path(dr2s$getOutdir(), "plot.sequence.png")
knitr::include_graphics(imgPath)
Sequence logos of the heterozygous positions used for clustering for each haplotype.
Some positions are not perfectly separated, but still sufficient.
imgPath <- file.path(dr2s$getOutdir(), "plot.association.png")
knitr::include_graphics(imgPath)
imgPath <- file.path(dr2s$getOutdir(), "plot.correlogram.png")
knitr::include_graphics(imgPath)
## We need to provide the limits for plotting the auto-threshold.
limits <- unlist(dr2s$getLimits())
dr2s$plotPartitionSummary(limits = limits)
Iterative mapping
The iterative mapping step is used to obtain haplotype-specific consensus scaffolds. From this step on, haplotypes are always treated separately.
It starts with a mapping of haplotype-specific reads from the clustering step against the sample-specific reference from the mapInit step.
A consensus is drawn from the mapping and this consensus is used as a reference for the subsequent re-mapping.
The default is set to two iterations of re-mapping.
## Iteratie mapping of long-reads to create a scaffold consensus
dr2s <- mapIter(dr2s)
Again, the coverage is plotted and remaining indels and heterozygous positions are highlighted:
imgPath <- file.path(dr2s$getOutdir(), "plot.mapIter.png")
knitr::include_graphics(imgPath)
## Collect the plots for each iteration and haplotype
plotlist <- purrr::map(seq_len(dr2s$getIterations()), function(iteration) {
iteration <- 1
suppressWarnings(dr2s$plotMapIterSummary(
thin = 0.1, width = 4, iteration = iteration, drop.indels = TRUE))
})
## arrange with cowplot
cowplot::plot_grid(plotlist = plotlist, nrow = dr2s$getIterations())
Assign short-reads to clusters
Short-reads are assigned to the derived clusters based on the mapping from the mapInit step. It is ensured that the distribution to the clusters is not too uneven.
## Cluster the short-reads
dr2s <- partitionShortreads(dr2s)
Polishing the scaffold consensus
The scaffold consensus derived from long-reads is now polished with the more accurate short-reads, i.e. the clustered short-reads are mapped against the consensus and a new, more accurate, consensus is drawn.
## Polish with short-reads and draw the final consensus
dr2s <- mapFinal(dr2s)
The coverage is again plotted for long- and short-reads of all haplotypes.
imgPath <- file.path(dr2s$getOutdir(), "plot.mapFinal.png")
knitr::include_graphics(imgPath)
## Plot coverage and base frequency
readtypes <- if (dr2s$hasShortreads()) c("LR", "SR") else "LR"
plotRows <- if (dr2s$hasShortreads()) 2 else 1
## readtype = "LR"
plotlist <- purrr::map(readtypes, function(readtype) {
suppressWarnings(dr2s$plotMapFinalSummary(readtype = readtype, thin = 0.25, width = 20))
})
p <- cowplot::plot_grid(plotlist = plotlist, nrow = plotRows, labels = readtypes, hjust = -0.25)
p
Remap and Report
The polished sequences may still not be perfect and need to be checked using the report function.
report does the following:
1. Possible problematic positions are reported to a csv file in the report subfolder.
2. The sequences are written to fasta file in the report subfolder.
3. The alignment of all haplotypes is written to a text file and as html in the report subfolder.
4. It is checked if the length of homopolymers is concordant between the consensus sequence and the mode value of short-reads.
5. Long- and short-reads are remapped to the final consensus for visualization with IGV.
## Report possible problems.
dr2s <- report(dr2s)
The coverage is plotted as it is in the mapFinal step. Additionally, the homopolymer distribution is plotted as a histogram of homopolymer lengths in all haplotypes.
imgPath <- file.path(dr2s$getOutdir(), "plot.remap.png")
knitr::include_graphics(imgPath)
## Plot coverage and base frequency
readtypes <- if (dr2s$hasShortreads()) c("LR", "SR") else "LR"
plotRows <- if (dr2s$hasShortreads()) 2 else 1
## readtype = "LR"
plotlist <- purrr::map(readtypes, function(readtype) {
suppressWarnings(dr2s$plotRemapSummary(readtype = readtype, thin = 0.25, width = 20))
})
p <- cowplot::plot_grid(plotlist = plotlist, nrow = plotRows, labels = readtypes, hjust = -0.25)
p
!! TODO show the homopolymer plot by loading it from file
Post-processing
You can now have a look at possible problems to resolve by looking at the file directly:
filepath <- file.path(dr2s$getOutdir(), "report", "problems.tsv")
file.edit(filepath)
Or you can load the tibble from the dr2s object:
knitr::kable(dr2s$consensus$variants)
The consensus sequences might now be manually edited. This is carried out in a text representation of the alignment of the haplotypes.
Positions which needs to be edited are best inferred by a visual inspection of the mapping using the IGV genome browser.
Positions that are still ambiguous are represented in the alignment by the respective IUPAC code, e.g. by 'K' for an 'A' or 'G'.
The alignment should be loaded into the editor using the checkAlignmentFile() command.
checkAlignmentFile() creates a copy of the original alignment file which can be edited.
The text file is by default loaded into the systems default editor.
Use the editor = "rstudio" parameter for loading the file into rstudio during an interactive session.
## Check the alignment file.
## If the file should be edited within rstudio, this must be supplied through the `editor` parameter.
## By default, the command will open your systems default editor.
checkAlignmentFile(dr2s, editor = "rstudio")
A pair-wise alignment of two haplotypes looks like this:
filepath <- list.files(
file.path(dr2s$getOutdir(), "report"),
pattern = "unchecked.psa$", full.names = TRUE)
cat(readLines(filepath)[1:38], sep = '\n')
It is recommended to remap the reads to the altered consensus sequence after manual edits.
The remapping is carried out by remapAndReport().
This function checks the alignment file for updated sequences and remaps reads to consensus sequences.
The coverage plot and possible problems are also updated.
## Remap reads to the updated consensus
remapAndReport(dr2s)
The consensus can be "checked out" if it looks correct in the coverage plot or IGV.
This can be done using the reportCheckedConsensus() function.
The function creates a "checked" subfolder in the output directory and writes the final sequences as fasta files together with the final alignment.
## Report the final consensus sequences
reportCheckedConsensus(dr2s)
Extended configuration
datadir: <path to the data>
outdir: <output directory>
pipeline: SR # SR: with short-reads, LR: only long-reads
format: yaml
longreads: # long-read configuration
type: pacbio # type of long-reads, pacbio or nanopore
dir: pacbio_lowQual # directory containing the fastq files. Either subdirectory of <datadir> or absolute path
shortreads: # short-read configuration
type: illumina # type of short-reads. Usually illumina
dir: illumina # directory containing the fastq files. Either subdirectory of <datadir> or absolute path
opts: # Options for the basic workflow steps
mapInit:
includeDeletions: yes # Include deletions for creating a first reference?
includeInsertions: yes # Include insertions for creating a first reference?
callInsertionThreshold: 0.18 # Call insertions from this fraction of insertions at a position
microsatellite: yes # Remap reads to its consensus to better infer repetetive regions
forceMapping: yes # force a mapping, even if a file with the current name already exists.
topx: no # Use only the "best matching" number of reads. Or the best matching reads using 'auto'
lowerLimit: 200.0 # Minimal number of reads picked with topx: "auto".
createIgv: yes # Create Igv config files and sample the mappings
partitionLongreads:
threshold: 0.1 # variant calling threshold
distAlleles: 2.0 # expected number of alleles (if known)
selectCorrelatedPositions: no # Use only informative variants (most useful when running only with long-reads)
selectAllelesBy: distance # automatically select found clusters by distance of a pwm or by count
minClusterSize: 20.0 # minimal number of reads to define a cluster
mapIter:
iterations: 2.0 # Number of iterative mappings
mapFinal:
includeDeletions: yes # Call deletions in the final consensus?
includeInsertions: yes # Call insertions in the final consensus?
callInsertionThreshold: 0.18 # Call insertions from this fraction of insertions at a position
report:
remap: yes # Remap the reads to the final consensus for inspection in IGV?
checkHpCount: yes # Compare homopolymer length of the consensus with the mode value in all short-reads
hpCount: 10.0 # minimal number of repeats to define a homopolymer
samples:
'1':
sampleId: sample1
locus: HLA-A
reference: HLA-A
distAlleles: 2
details:
second: another
third: .na
LIMS_DONOR_ID: .na
Spendernummer: .na
notes: ~
'2':
sampleId: sample2
locus: HLA-A
reference: HLA-A
distAlleles: 2
details:
second: another
third: .na
LIMS_DONOR_ID: .na
Spendernummer: .na
notes: ~
DKMS-LSL/dr2s documentation built on March 14, 2021, 2:46 p.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
DR2S is a tool to create fully-phased reference grade sequences of heterozygous loci. It is especially designed for working with the highly polymorphic HLA and KIR genes. A detailed description of the algorithm and how it works can be found at bioRxiv.
Input:
The usual inputs to DR2S are fastqs of two sequencing experiments, one of short-reads of shotgun-sequencing (e.g. Illumina), the second of long-reads from third-generation sequencing, e.g. ONT Nanopore or PacBio. Additionally, a generic reference needs to be give, either as an HLA locus, HLA or KIR allele identifier or path to a fasta file.
Config
Input parameters and paths are either bundled in a config file (yaml or json) or can be provided in an interactive R session.
Possible arguments are described in the function documentation of DR2S::createDR2SConf() ?DR2S::createDR2SConf().
A default config would look like this:
datadir: <path to the data> outdir: <output directory> pipeline: SR # SR: with short-reads, LR: only long-reads format: yaml longreads: # long-read configuration type: pacbio # type of long-reads, pacbio or nanopore dir: pacbio_lowQual # directory containing the fastq files. Either subdirectory of <datadir> or absolute path shortreads: # short-read configuration type: illumina # type of short-reads. Usually illumina dir: illumina # directory containing the fastq files. Either subdirectory of <datadir> or absolute path sampleId: sample1 locus: HLA-A reference: HLA-A # Either an HLA or KIR allele identifier, an HLA locus or the path to a reference in fasta format details: # Any field in details will go to the header of the final fasta. partnerAllele: another notes: xx
The yaml config can be read using the readDR2SConf() function.
# library(DR2S) devtools::load_all() configFile <- system.file("test/dr2s_conf.yaml", package = "DR2S") # read the example config config <- readDR2SConf(configFile)
The configuration can also be provided via R in an interactive session or script:
library(DR2S) config <- createDR2SConf( sample = "sample3", locus = "HLA-A", longreads = list(dir = "nanopore_sampled", type = "nanopore", mapper = "minimap"), shortreads = list(dir = "Illumina_sampled", type = "illumina", mapper = "rsubread"), datadir = system.file("testData", package = "DR2S"), outdir = tempdir(), reference = "HLA-A" )
Initialization
The DR2S object must be initialized with a valid config using the InitDR2S() command.
## Initialize the DR2S object dr2s <- InitDR2S(config) ## The output directory can be cleaned if there remains data from previous attempts using the clear() function. This deletes all the data and re-initialize the project. clear(dr2s) ## Look what it can tell you print(dr2s)
InitDR2S() creates the output directory, writes the config as a json file, creates the log and copies the reference.
By default, a folder structure will be created for the output directory consisting of the outdir param, the locus and the sampleId. This can be suppressed by setting createOutdir=FALSE .
The DR2S object is an R6 class, which means it behaves like a reference class. It is always modified in-place and an assignment after the initialization is not necessary.
All methods in the DR2S workflow expect a DR2S object as input and return a DR2S object, which means they can be chained using magrittr's pipe operator %>%.
Initial mapping
The mapInit step creates a first sample-specific consensus sequence by a mapping of the short-reads to the generic reference and subsequent consensus calling. This sequence is still heterozygous. The long- and short-reads are mapped to this sample-specific consensus. Options are drawn from the initial config stored in the DR2S object.
## Run mapInit dr2s <- mapInit(dr2s)
The mapping or coverage plot is created for long- and short-reads and is stored in the output directory.
imgPath <- file.path(dr2s$getOutdir(), "plot.mapInit.png") knitr::include_graphics(imgPath)
cluster the long-reads
The clustering of reads based on the long-reads is carried out using the partitionLongreds command.
Heterozygous positions are derived from short-reads of the mapIninit step. Long-reads are clustered based on these heterozygous positions.
## cluster the long-reads dr2s <- partitionLongreads(dr2s)
Different plots are created to visualize the clustering.
imgPath <- file.path(dr2s$getOutdir(), "plot.partition.png") knitr::include_graphics(imgPath)
The top panel shows a histogram of the haplotype membership coefficient for assigned haplotypes. Different colors indicate different haplotypes. The middle panel shows the tree of hierarchical clustering. The blue line indicates the cut-height for haplotype inference. Leafs are single long-reads. The bottom panel shows assignment of reads assigned to each cluster to all other clusters. This is only useful in cases of more than two haplotypes.
imgPath <- file.path(dr2s$getOutdir(), "plot.sequence.png") knitr::include_graphics(imgPath)
Sequence logos of the heterozygous positions used for clustering for each haplotype. Some positions are not perfectly separated, but still sufficient.
imgPath <- file.path(dr2s$getOutdir(), "plot.association.png") knitr::include_graphics(imgPath)
imgPath <- file.path(dr2s$getOutdir(), "plot.correlogram.png") knitr::include_graphics(imgPath)
## We need to provide the limits for plotting the auto-threshold. limits <- unlist(dr2s$getLimits()) dr2s$plotPartitionSummary(limits = limits)
Iterative mapping
The iterative mapping step is used to obtain haplotype-specific consensus scaffolds. From this step on, haplotypes are always treated separately. It starts with a mapping of haplotype-specific reads from the clustering step against the sample-specific reference from the mapInit step. A consensus is drawn from the mapping and this consensus is used as a reference for the subsequent re-mapping. The default is set to two iterations of re-mapping.
## Iteratie mapping of long-reads to create a scaffold consensus dr2s <- mapIter(dr2s)
Again, the coverage is plotted and remaining indels and heterozygous positions are highlighted:
imgPath <- file.path(dr2s$getOutdir(), "plot.mapIter.png") knitr::include_graphics(imgPath)
## Collect the plots for each iteration and haplotype plotlist <- purrr::map(seq_len(dr2s$getIterations()), function(iteration) { iteration <- 1 suppressWarnings(dr2s$plotMapIterSummary( thin = 0.1, width = 4, iteration = iteration, drop.indels = TRUE)) }) ## arrange with cowplot cowplot::plot_grid(plotlist = plotlist, nrow = dr2s$getIterations())
Assign short-reads to clusters
Short-reads are assigned to the derived clusters based on the mapping from the mapInit step. It is ensured that the distribution to the clusters is not too uneven.
## Cluster the short-reads dr2s <- partitionShortreads(dr2s)
Polishing the scaffold consensus
The scaffold consensus derived from long-reads is now polished with the more accurate short-reads, i.e. the clustered short-reads are mapped against the consensus and a new, more accurate, consensus is drawn.
## Polish with short-reads and draw the final consensus dr2s <- mapFinal(dr2s)
The coverage is again plotted for long- and short-reads of all haplotypes.
imgPath <- file.path(dr2s$getOutdir(), "plot.mapFinal.png") knitr::include_graphics(imgPath)
## Plot coverage and base frequency readtypes <- if (dr2s$hasShortreads()) c("LR", "SR") else "LR" plotRows <- if (dr2s$hasShortreads()) 2 else 1 ## readtype = "LR" plotlist <- purrr::map(readtypes, function(readtype) { suppressWarnings(dr2s$plotMapFinalSummary(readtype = readtype, thin = 0.25, width = 20)) }) p <- cowplot::plot_grid(plotlist = plotlist, nrow = plotRows, labels = readtypes, hjust = -0.25) p
Remap and Report
The polished sequences may still not be perfect and need to be checked using the report function.
report does the following:
1. Possible problematic positions are reported to a csv file in the report subfolder.
2. The sequences are written to fasta file in the report subfolder.
3. The alignment of all haplotypes is written to a text file and as html in the report subfolder.
4. It is checked if the length of homopolymers is concordant between the consensus sequence and the mode value of short-reads.
5. Long- and short-reads are remapped to the final consensus for visualization with IGV.
## Report possible problems. dr2s <- report(dr2s)
The coverage is plotted as it is in the mapFinal step. Additionally, the homopolymer distribution is plotted as a histogram of homopolymer lengths in all haplotypes.
imgPath <- file.path(dr2s$getOutdir(), "plot.remap.png") knitr::include_graphics(imgPath)
## Plot coverage and base frequency readtypes <- if (dr2s$hasShortreads()) c("LR", "SR") else "LR" plotRows <- if (dr2s$hasShortreads()) 2 else 1 ## readtype = "LR" plotlist <- purrr::map(readtypes, function(readtype) { suppressWarnings(dr2s$plotRemapSummary(readtype = readtype, thin = 0.25, width = 20)) }) p <- cowplot::plot_grid(plotlist = plotlist, nrow = plotRows, labels = readtypes, hjust = -0.25) p
!! TODO show the homopolymer plot by loading it from file
Post-processing
You can now have a look at possible problems to resolve by looking at the file directly:
filepath <- file.path(dr2s$getOutdir(), "report", "problems.tsv") file.edit(filepath)
Or you can load the tibble from the dr2s object:
knitr::kable(dr2s$consensus$variants)
The consensus sequences might now be manually edited. This is carried out in a text representation of the alignment of the haplotypes.
Positions which needs to be edited are best inferred by a visual inspection of the mapping using the IGV genome browser.
Positions that are still ambiguous are represented in the alignment by the respective IUPAC code, e.g. by 'K' for an 'A' or 'G'.
The alignment should be loaded into the editor using the checkAlignmentFile() command.
checkAlignmentFile() creates a copy of the original alignment file which can be edited.
The text file is by default loaded into the systems default editor.
Use the editor = "rstudio" parameter for loading the file into rstudio during an interactive session.
## Check the alignment file. ## If the file should be edited within rstudio, this must be supplied through the `editor` parameter. ## By default, the command will open your systems default editor. checkAlignmentFile(dr2s, editor = "rstudio")
A pair-wise alignment of two haplotypes looks like this:
filepath <- list.files( file.path(dr2s$getOutdir(), "report"), pattern = "unchecked.psa$", full.names = TRUE) cat(readLines(filepath)[1:38], sep = '\n')
It is recommended to remap the reads to the altered consensus sequence after manual edits.
The remapping is carried out by remapAndReport().
This function checks the alignment file for updated sequences and remaps reads to consensus sequences.
The coverage plot and possible problems are also updated.
## Remap reads to the updated consensus remapAndReport(dr2s)
The consensus can be "checked out" if it looks correct in the coverage plot or IGV.
This can be done using the reportCheckedConsensus() function.
The function creates a "checked" subfolder in the output directory and writes the final sequences as fasta files together with the final alignment.
## Report the final consensus sequences reportCheckedConsensus(dr2s)
Extended configuration
datadir: <path to the data> outdir: <output directory> pipeline: SR # SR: with short-reads, LR: only long-reads format: yaml longreads: # long-read configuration type: pacbio # type of long-reads, pacbio or nanopore dir: pacbio_lowQual # directory containing the fastq files. Either subdirectory of <datadir> or absolute path shortreads: # short-read configuration type: illumina # type of short-reads. Usually illumina dir: illumina # directory containing the fastq files. Either subdirectory of <datadir> or absolute path opts: # Options for the basic workflow steps mapInit: includeDeletions: yes # Include deletions for creating a first reference? includeInsertions: yes # Include insertions for creating a first reference? callInsertionThreshold: 0.18 # Call insertions from this fraction of insertions at a position microsatellite: yes # Remap reads to its consensus to better infer repetetive regions forceMapping: yes # force a mapping, even if a file with the current name already exists. topx: no # Use only the "best matching" number of reads. Or the best matching reads using 'auto' lowerLimit: 200.0 # Minimal number of reads picked with topx: "auto". createIgv: yes # Create Igv config files and sample the mappings partitionLongreads: threshold: 0.1 # variant calling threshold distAlleles: 2.0 # expected number of alleles (if known) selectCorrelatedPositions: no # Use only informative variants (most useful when running only with long-reads) selectAllelesBy: distance # automatically select found clusters by distance of a pwm or by count minClusterSize: 20.0 # minimal number of reads to define a cluster mapIter: iterations: 2.0 # Number of iterative mappings mapFinal: includeDeletions: yes # Call deletions in the final consensus? includeInsertions: yes # Call insertions in the final consensus? callInsertionThreshold: 0.18 # Call insertions from this fraction of insertions at a position report: remap: yes # Remap the reads to the final consensus for inspection in IGV? checkHpCount: yes # Compare homopolymer length of the consensus with the mode value in all short-reads hpCount: 10.0 # minimal number of repeats to define a homopolymer samples: '1': sampleId: sample1 locus: HLA-A reference: HLA-A distAlleles: 2 details: second: another third: .na LIMS_DONOR_ID: .na Spendernummer: .na notes: ~ '2': sampleId: sample2 locus: HLA-A reference: HLA-A distAlleles: 2 details: second: another third: .na LIMS_DONOR_ID: .na Spendernummer: .na notes: ~
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.