Package 'GenomicScores'

Title: Infrastructure to work with genomewide position-specific scores
Description: Provide infrastructure to store and access genomewide position-specific scores within R and Bioconductor.
Authors: Robert Castelo [aut, cre], Pau Puigdevall [ctb], Pablo Rodríguez [ctb]
Maintainer: Robert Castelo <[email protected]>
License: Artistic-2.0
Version: 2.19.0
Built: 2024-11-18 03:47:51 UTC
Source: https://github.com/bioc/GenomicScores

Help Index

Exploring genomic scores resources

Description

Functions to explore genomic scores resources.

Usage

availableGScores(use.internet=FALSE)
getGScores(x, accept.license=FALSE)

Arguments

x

Acharacter vector of length 1 specifiying the genomic scores resource to fetch.
use.internet

A logical value specifying whether we want to check through the internet whether the annotation packages and AnnotationHub resources are ready to be downloaded. By default, i.e. use.internet=FALSE, this is not checked to speed up the execution of this function.
accept.license

A logical value specifying whether we accept the license under which we can use the requested data, when the data is distributed under such a license.

Details

The function availableGScores() shows genomic score sets available as AnnotationHub online resources.

Value

The function availableGScores() returns a data.frame object with a row for each available resource of genomic scores and the following four columns:

  • Name: Name of the Bioconductor package or AnnotationHub resource.

  • Organism: Organism on which the genomic scores are defined.

  • Category: Category to which the genomic scores belong to.

  • Installed: Whether the resource is installed as a package (TRUE) or not (FALSE).

  • Cached: Whether the resource is available within the local cache of the AnnotationHub (TRUE) or not (FALSE).

  • BiocManagerInstall: Whether the resource can be installed as an annotation package through BiocManager::install() (TRUE) or not (FALSE).

  • AnnotationHub: Whether the resource can be downloaded as a GScores object through the AnnotationHub, using the function getGScores() (TRUE), or not (FALSE).

The function getGScores() returns a GScores object.

Author(s)

R. Castelo

References

Puigdevall, P. and Castelo, R. GenomicScores: seamless access to genomewide position-specific scores from R and Bioconductor. Bioinformatics, 18:3208-3210, 2018.

See Also

getGScores() phastCons100way.UCSC.hg38 MafDb.1Kgenomes.phase1.hs37d5

Examples

availableGScores()

## Not run: 
gsco <- getGScores("cadd.v1.6.hg38")

## End(Not run)

Accessing genomic gscores

Description

Functions to access genomic gscores through GScores objects.

Usage

## S4 method for signature 'GScores,GenomicRanges'
gscores(x, ranges, ...)
## S4 method for signature 'GScores,character'
gscores(x, ranges, ...)
## S4 method for signature 'GScores'
score(x, ..., simplify=TRUE)

Arguments

x

A GScores object.
ranges

A GenomicRanges object with positions from where to retrieve genomic scores, or a character string vector with identifiers associated by the data producer with the genomic scores, e.g., dbSNP 'rs' identifiers in the case of some MafDb.* packages.
...

In the call to the gscores() method one can additionally specify the following optional parameters:

  • pop: Character string vector specifying the scores populations to query, when there is more than one. By default, its value is 'defaultPopulation(x)'. Use populations() to find out the available scores populations.

  • type: Character string specifying the type of genomic position being sought, which can be a single nucleotide range (snr), by default, or a nonsnr spanning multiple nucleotides. The latter is the case of indel variants in minor allele frequency data.

  • scores.only: Flag setting whether only the scores should be returned as a numeric vector (TRUE), instead of returning them as a metadata column in a GRanges object (FALSE, default).

  • summaryFun: Function to summarize genomic scores when more than one position is retrieved. By default, this is set to the arithmetic mean, i.e., the mean() function.

  • quantized: Flag setting whether the genomic scores should be returned quantized (TRUE) or dequantized (FALSE, default).

  • ref: Vector of reference alleles in the form of either a character vector, a DNAStringSet object or a DNAStringSetList object. This argument is used only when either there are multiple scores per position or x is a MafDb.* package.

  • alt: Vector of alternative alleles in the form of either a character vector, a DNAStringSet object or a DNAStringSetList object. This argument is used only when either there are multiple scores per position or x is a MafDb.* package.

  • minoverlap: Integer value passed internally to the function findOverlaps() from the IRanges package, when querying genomic positions associated with multiple-nucleotide ranges (nonSNRs). By default, minoverlap=1L, which assumes that the sought nonSNRs are stored as in VCF files, using the nucleotide composition of the reference sequence. This argument is only relevant for genomic scores associated with nonSNRs.

  • caching: Flag setting whether genomic scores per chromosome should be kept cached in memory (TRUE, default) or not (FALSE). The latter option minimizes the memory footprint but slows down the performance when the gscores() method is called multiple times.

simplify

Flag setting whether the result should be simplified to a vector (TRUE, default) if possible. This happens when scores from a single population are queried.

Details

The method gscores() takes as first argument a GScores object, previouly loaded from either an annotation package or an AnnotationHub resource; see getGScores().

The arguments ref and alt serve two purposes. One, when there are multiple scores per position, such as with CADD or M-CAP, and we want to select a score matching a specific combination of reference and alternate alleles. The other purpose is when the GScores object x is a MafDb.* package, then by providing ref and alt alelles we will get separate frequencies for reference and alternate alleles. The current lossy compression of these values yields a correct assignment for biallelic variants in the corresponding MafDb.* package and an approximation for multiallelic ones.

Value

The method gscores() returns a GRanges object with the genomic scores in a metadata column called score. The method score() returns a numeric vector with the genomic scores.

Author(s)

R. Castelo

References

Puigdevall, P. and Castelo, R. GenomicScores: seamless access to genomewide position-specific scores from R and Bioconductor. Bioinformatics, 18:3208-3210, 2018.

See Also

phastCons100way.UCSC.hg38 MafDb.1Kgenomes.phase1.hs37d5

Examples

## one genomic range of width 5
gr1 <- GRanges(seqnames="chr22", IRanges(start=50528591, width=5))
gr1

## five genomic ranges of width 1
gr2 <- GRanges(seqnames="chr22", IRanges(start=50528591:50528596, width=1))
gr2

## accessing genomic gscores from an annotation package
if (require(phastCons100way.UCSC.hg38)) {
  library(GenomicRanges)

  gsco <- phastCons100way.UCSC.hg38
  gsco
  gscores(gsco, gr1)
  score(gsco, gr1)
  gscores(gsco, gr2)
  populations(gsco)
  gscores(gsco, gr2, pop="DP2")
}

if (require(MafDb.1Kgenomes.phase1.hs37d5)) {
  mafdb <- MafDb.1Kgenomes.phase1.hs37d5
  mafdb
  populations(mafdb)

  ## lookup allele frequencies for SNP rs1129038, located at 15:28356859, a
  ## SNP associated to blue and brown eye colors as reported by Eiberg et al.
  ## Blue eye color in humans may be caused by a perfectly associated founder
  ## mutation in a regulatory element located within the HERC2 gene
  ## inhibiting OCA2 expression. Human Genetics, 123(2):177-87, 2008
  ## [http://www.ncbi.nlm.nih.gov/pubmed/18172690]
  gscores(mafdb, GRanges("15:28356859"), pop=populations(mafdb))
  gscores(mafdb, "rs1129038", pop=populations(mafdb))
}

GScores objects

Description

The goal of the GenomicScores package is to provide support to store and retrieve genomic scores associated to physical nucleotide positions along a genome. This is achieved through the GScores class of objects, which is a container for genomic score values.

Details

The GScores class attempts to provide a compact storage and efficient retrieval of genomic score values that have been typically processed and stored using some form of lossy compression. This class is currently based on a former version of the SNPlocs class defined in the BSgenome package, with the following slots:

provider

(character), the data provider such as UCSC.

provider_version

(character), the version of the data as given by the data provider, typically a date in some compact format.

download_url

(character), the URL of the data provider from where the original data were downloaded.

download_date

(character), the date on which the data were downloaded.

reference_genome

(GenomeDescription), object with information about the reference genome whose physical positions have the genomic scores.

data_pkgname

(character), name given to the set of genomic scores associated to a particular genome. When the genomic scores are stored within an annotation package, then this corresponds to the name of that package.

data_dirpath

(character), absolute path to the local directory where the genomic scores are stored in one file per genome sequence.

data_serialized_objnames

(character), named vector of filenames pointing to files containing the genomic scores in one file per genome sequence. The names of this vector correspond to the genome sequence names.

data_group

(character), name denoting a category of genomic scores to which the scores stored in the object belong to. Typical values are "Conservation", "MAF", "Pathogenicity", etc.

data_tag

(character), name identifying the genomic scores stored in the object and which can be used, for instance, to assign a column name storing these scores.

data_pops

(character), vector of character strings storing score population names. The term "default" is reserved to denote a score set that is not associated to a particular population name and is used by default.

data_nonsnrs

(logical), flag indicating whether the object stores genomic scores associated with non-single nucleotide ranges.

data_nsites

(integer), number of sites in the genome associated with the genomic scores stored in the object.

.data_cache

(environment), data structure where objects storing genomic scores are cached into main memory.

The goal of the design behind the GScores class is to load into main memory only the objects associated with the queried sequences to minimize the memory footprint, which may be advantageous in workflows that parallelize the access to genomic scores by genome sequence.

GScores objects are created either from AnnotationHub resources or when loading specific annotation packages that store genomic score values. Two such annotation packages are:

phastCons100way.UCSC.hg19

Nucleotide-level phastCons conservation scores from the UCSC Genome Browser calculated from multiple genome alignments from the human genome version hg19 to 99 vertebrate species.

phastCons100way.UCSC.hg38

Nucleotide-level phastCons conservation scores from the UCSC Genome Browser calculated from multiple genome alignments from the human genome version hg38 to 99 vertebrate species.

Constructor

GScores(provider, provider_version, download_url, download_date, reference_genome, data_pkgname, data_dirpath, data_serialized_objnames, default_pop, data_tag):

Creates a GScores object. In principle, the end-user needs not to call this function.

provider

character string, containing the data provider.

provider_version

character string, containing the version of the data as given by the data provider.

download_url

character string, containing the URL of the data provider from where the original data were downloaded.

reference_genome

GenomeDescription, storing the information about the associated reference genome.

data_pkgname

character string, name given to the set of genomic scores stored through this object.

data_dirpath

character string, absolute path to the local directory where the genomic scores are stored.

data_serialized_objname

character string vector, containing filenames where the genomic scores are stored.

default_pop

character string, containing the name of the default scores population.

data_group

character string, containing a name that indicates a category of genomic scores to which the scores in the object belong to. Typical names could be "Conservation", "MAF", etc.

data_tag

character string, containing a tag that succintly labels genomic scores from a particular source. This can be used to automatically give, for instance, a name to a column storing genomic scores in data frame object. Its default value takes the prefix of the package name.

Accessors

name(x):

get the name of the set of genomic scores.

type(x):

get the substring of the name of the set of genomic scores comprised between the first character until the first period. This should typically match the type of genomic scores such as, phastCons, phyloP, etc.

provider(x):

get the data provider.

providerVersion(x):

get the provider version.

organism(x):

get the organism associated with the genomic scores.

seqlevelsStyle(x):

get the genome sequence style.

seqinfo(x):

get the genome sequence information.

seqnames(x):

get the genome sequence names.

seqlengths(x):

get the genome sequence lengths.

populations(x):

get the identifiers of the available scores populations. If only one scores population is available, then it shows only the term default.

defaultPopulation(x):

get or set the default population of scores.

gscoresCategory(x):

get or set the genomic scores category label.

gscoresTag(x):

get or set the genomic scores tag label.

gscoresNonSNRs(x):

get whether there are genomic scores associated with non-single nucleotide ranges.

nsites(x):

get the number of sites in the genome with genomic scores.

qfun(x):

get the quantizer function.

dqfun(x):

get the dequantizer function.

citation(x):

get citation information for the genomic scores data in the form of a bibentry object.

Author(s)

R. Castelo

References

Puigdevall, P. and Castelo, R. GenomicScores: seamless access to genomewide position-specific scores from R and Bioconductor. Bioinformatics, 18:3208-3210, 2018.

See Also

gscores() score() phastCons100way.UCSC.hg38

Examples

## one genomic range of width 5
gr1 <- GRanges(seqnames="chr22", IRanges(start=50528591, width=5))
gr1

## five genomic ranges of width 1
gr2 <- GRanges(seqnames="chr22", IRanges(start=50528591:50528596, width=1))
gr2

## supporting annotation packages with genomic scores
if (require(phastCons100way.UCSC.hg38)) {
  library(GenomicRanges)

  phast <- phastCons100way.UCSC.hg38
  phast
  gscores(phast, gr1)
  score(phast, gr1)
  gscores(phast, gr2)
  populations(phast)
  gscores(phast, gr2, pop="DP2")
}

## supporting AnnotationHub resources
## Not run: 
availableGScores()
phast <- getGScores("phastCons100way.UCSC.hg38")
phast
gscores(phast, gr1)

## End(Not run)

## metadata from a GScores object
name(phast)
type(phast)
provider(phast)
providerVersion(phast)
organism(phast)
seqlevelsStyle(phast)
seqinfo(phast)
head(seqnames(phast))
head(seqlengths(phast))
gscoresTag(phast)
populations(phast)
defaultPopulation(phast)
qfun(phast)
dqfun(phast)
citation(phast)

GenomicScores shiny app

Description

Starts an interactive GenomicScores shiny web app.

Usage

igscores()

Details

The goal of the GenomicScores package is to provide support to store and retrieve genomic scores associated to physical nucleotide positions along a genome.

The igscores() function starts an interactive shiny web app that allows the user to query annotation packages storing genomic scores. Internally, it calls to the function gscores(); see its manual page for a description of the arguments and their default and alternative values.

Value

None.

Author(s)

P. Rodríguez and R. Castelo

References

Puigdevall, P. and Castelo, R. GenomicScores: seamless access to genomewide position-specific scores from R and Bioconductor. Bioinformatics, 18:3208-3210, 2018.

See Also

gscores

Examples

## Not run: 
igscores() ## this will open your browser with the GenomicScores shiny web app

## End(Not run)

Building genomic gscores packages

Description

Build a genomic gscores packages from a GScores object.

Usage

makeGScoresPackage(gsco, version, maintainer, author,
                   destDir=".", license="Artistic-2.0")

Arguments

gsco

GScores. The GScores object from which the package will be created.
version

Character. Version of the package.
maintainer

Character. Maintainer of the package, including email.
author

Character. Author of the package.
destDir

Character. The path to the directory where the package will be created.
license

Character. The license of the package.

Details

This function allows one to create an R package from a GScores object. This may be useful if one wants to have a tar-ball package version of genomic scores available only through the AnnotationHub; see the vignette.

Value

It returns invisibily the package directory.

Author(s)

R. Castelo

References

Puigdevall, P. and Castelo, R. GenomicScores: seamless access to genomewide position-specific scores from R and Bioconductor. Bioinformatics, 18:3208-3210, 2018.

See Also

availableGScores() getGScores()

Examples

## accessing genomic scores from AnnotationHub resources
## and building a package from them
## Not run: 
availableGScores()
gsco <- getGScores("fitCons.UCSC.hg19")
makeGScoresPackage(gsco, version="1.0", maintainer="me <[email protected]>", author="me")

## End(Not run)

Sampling genomic gscores

Description

Function for randomly sampling genomic gscores from GScores objects.

Usage

## S4 method for signature 'GScores,missing'
rgscores(n, object, ...)
## S4 method for signature 'missing,GScores'
rgscores(n, object, ...)
## S4 method for signature 'numeric,GScores'
rgscores(n, object, ...)
## S4 method for signature 'integer,GScores'
rgscores(n, object, ...)

Arguments

n

Number of scores to sample.
object

A GScores object.
...

In the call to the rgscores() method one can additionally set the following arguments:

  • pop Character string vector of length one, specifying the scores population from which we want to sample scores. By default, its value is defaultPopulation(object). Use populations() to find out the available scores populations.

  • scores.only Flag setting whether only the scores should be returned as a numeric vector (TRUE), instead of returning them as a metadata column in a GRanges object (FALSE, default).

  • ranges Either a GRanges object or a character string vector of sequence names. Scores will be sampled from the given genomic regions, which by default correspond to the entire genomic space of object. Currently, only entire chromosomes are considered.

Details

The method rgscores() samples scores randomly from a GScores object.

Value

A GRanges object with the sampled genomic positions and scores. When scores.only=TRUE then a numeric vector is returned with the sampled scores.

Author(s)

R. Castelo

References

Puigdevall, P. and Castelo, R. GenomicScores: seamless access to genomewide position-specific scores from R and Bioconductor. Bioinformatics, 18:3208-3210, 2018.

See Also

phastCons100way.UCSC.hg38 MafDb.1Kgenomes.phase1.hs37d5

Examples

## accessing genomic gscores from an annotation package
if (require(phastCons100way.UCSC.hg38)) {
  library(GenomicRanges)

  phast <- phastCons100way.UCSC.hg38
  set.seed(123)
  rgscores(10L, phast, ranges=c("chr22", "chrY"))
}

if (require(MafDb.1Kgenomes.phase1.hs37d5)) {
  mafdb <- MafDb.1Kgenomes.phase1.hs37d5
  set.seed(123)
  rgscores(10L, mafdb, ranges=c("21", "22"))
}

Discovering genomic gscores

Description

Functions to discover which genomic scores are present in given genomic ranges through GScores objects.

Usage

## S4 method for signature 'GScores,GenomicRanges'
wgscores(x, ranges, ...)

Arguments

x

A GScores object.
ranges

A GenomicRanges object with genomic ranges where to search for genomic scores.
...

In the call to the wgscores() method one can additionally set the following arguments:

  • pop: Character string vector specifying the scores populations to query, when there is more than one. By default, its value is 'defaultPopulation(x)'. Use populations() to find out the available scores populations.

  • type: Character string specifying the type of genomic position being sought, which can be a single nucleotide range (snr), by default, or a nonsnr spanning multiple nucleotides. The latter is the case of indel variants in minor allele frequency data.

  • caching: Flag setting whether genomic scores per chromosome should be kept cached in memory (TRUE, default) or not (FALSE). The latter option minimizes the memory footprint but slows down the performance when the wgscores() method is called multiple times.

Details

The method wgscores() takes as first argument a GScores object, previouly loaded from either an annotation package or an AnnotationHub resource; see getGScores() and a GenomicRanges object as a second argument. It will search for which genomic scores fall within the provided genomic ranges and return them in an ordered GenomicRanges object with the scores in the metadata columns.

Value

The method wgscores() returns a GRanges object with the genomic scores in metadata columns named after the corresponding score population name.

Author(s)

R. Castelo

References

Puigdevall, P. and Castelo, R. GenomicScores: seamless access to genomewide position-specific scores from R and Bioconductor. Bioinformatics, 18:3208-3210, 2018.

See Also

MafDb.1Kgenomes.phase1.hs37d5

Examples

if (require(MafDb.1Kgenomes.phase1.hs37d5)) {
  mafdb <- MafDb.1Kgenomes.phase1.hs37d5
  mafdb
  populations(mafdb)

  ## lookup allele frequencies for SNP rs1129038, located at 15:28356859 in
  ## GRCh37, a SNP associated to blue and brown eye colors as reported by
  ## Eiberg et al. (Human Genetics, 2008; http://www.ncbi.nlm.nih.gov/pubmed/18172690).
  ## Blue eye color in humans may be caused by a perfectly associated founder
  ## mutation in a regulatory element located within the HERC2 gene inhibiting
  ## OCA2 expression.
  ##
  ## for the sake of illustrating this functionality let's create a
  ## GenomicRanges object with the SNP rs1129038 and enlarge its range
  ## by 200nt at each flank.
  snp <- GRanges("15:28356859")
  rngsnp <- flank(snp, width=100, both=TRUE)
  width(rngsnp)

  ## now use this genomic range to search for the rs1129038 SNP
  wgscores(mafdb, rngsnp)

  ## let's illustrate this same functionality for INDELs, starting
  ## from the specific INDEL rs113993960 that leads to the loss of
  ## phenylalanine at amino acid position 508 of the CFTR protein,
  ## commonly referred to as F508del in the CFTR gene, which is
  ## concretely a deletion of four nucleotides at position
  ## 7:117199644 in GRCh37 and enlarge its range by 20nt on each flank.
  indel <- GRanges(seqnames="chr7",
                   ranges=IRanges(start=117199644, width=4))
  rngindel <- flank(indel, width=20, both=TRUE)
  width(rngindel)

  ## now use this genomic range to search for the rs113993960 INDEL
  wgscores(mafdb, rngindel, type="nonsnrs")
}