Package 'bbmix'

The 'bbmix' package.

Description

Bayesian Beta-Binomial mixture model for RNA-seq genotyping

References

Stan Development Team (2018). RStan: the R interface to Stan. R package version 2.18.2. https://mc-stan.org

---

Call genotypes using beta binomial after model training

Description

Call genotypes using beta binomial after model training

Usage

call_gt(
  allele_counts_f,
  depth = 10,
  stan_f = NULL,
  legend_f,
  pop = "EUR",
  prob = 0.99,
  fisher_f = NULL,
  fisher = 30,
  cluster_f = NULL,
  out
)

Arguments

allele_counts_f	vector with file names with allele counts for SNPs
depth	min read count to call variant
stan_f	full name to stan object with model fit to extract mean of parameters. Defaults to the model trained with genome in a bottle reads. Otherwise this object can be generated with fit_bb function.
legend_f	full name for file with SNP info to get allele frequency for prior
pop	population to select AF for GT prior, defaults to EUR
prob	cut-off for making hard calls, defaults to 0.99
fisher_f	file with Fisher test to detect strand bias
fisher	cut_off for Fisher test to detect strand bias
cluster_f	file with info about SNP clusters
out	character with file name to save genotype output

Value

data table with genotype probabilities

Examples

## Retrive input files for running call_gt
counts_f <- system.file("extdata/input", "NA12878.chr22.Q20.allelicCounts.txt",
package = "bbmix",
mustWork = TRUE)

legend <- system.file("extdata/input", "1000GP_Phase3_chr22.legend",
package = "bbmix", mustWork = TRUE)

fisher_f <- system.file("extdata/input", "chr22.FS.Q20.alleleCounts.txt",
package = "bbmix", mustWork = TRUE)

cluster_f <- system.file("extdata/input", "fSNPs_22_RP_maf0_01_cluster3window35.txt",
package = "bbmix", mustWork = TRUE)

out <- paste0(tempdir() , "/NA12878.chrom22.gt.txt")

## Run call_gt:
call_gt(allele_counts_f = counts_f,
legend_f = legend,
fisher_f = fisher_f,
cluster_f = cluster_f,
out = out)

unlink(out)

---

call gt helper, calculate mean dbetabinom from all posterior samples

Description

call gt helper, calculate mean dbetabinom from all posterior samples

Usage

call_help(n, m, mu, lambda)

Arguments

n	counts alt allele
m	total counts
mu	vector with posterior draws for mu param
lambda	vector with posterior draws for lambda param

Value

mean of dbetabinom

---

Exclude fSNPs with no alternative allele in any sample. Also exclude fSNPs if all samples are hom.

Description

Exclude fSNPs with no alternative allele in any sample. Also exclude fSNPs if all samples are hom.

Usage

ex_alt_hom(gt_f, out)

Arguments

gt_f	character vector with file names with genotype calls per sample
out	file name to save output

Value

save file

Examples

gt_f <- system.file("extdata/output", "gt.NA12878.chr22.txt",
package = "bbmix",
mustWork = TRUE)
out <- tempfile()

## Running function
ex_alt_hom(gt_f, out)

unlink(out)

---

Fit beta binomial distribution to allelic counts for homozygous reference, heterozygous, homozygous alternative

Description

Fit beta binomial distribution to allelic counts for homozygous reference, heterozygous, homozygous alternative

Usage

fit_bb(
  counts_f,
  depth = 10,
  N = 1000,
  prefix = NULL,
  k = 3,
  alpha_p = c(1, 10, 499),
  beta_p = c(499, 10, 1),
  out,
  mc.cores = NULL
)

Arguments

counts_f	file name with allele counts for SNPs
depth	depth cut-off to use to select SNPs to fit distributions
N	number of SNPs to use for fitting
prefix	charcter with prefix to add for saving files, defaults to NULL
k	number of components for mixture model, defaults to 3
alpha_p	alpha parameter for the k components of alpha parameter
beta_p	beta paramenter for the k components of Beta parameter
out	character with dir name to save output
mc.cores	number of cores to use, defaults to parallel detected cores

Value

saves stan object to file

Examples

## Not run: 
## Retrive input files for running call_gt
counts_f <- system.file("extdata/input", "NA12878.chr22.Q20.allelicCounts.txt",
package = "bbmix",
mustWork = TRUE)

out <- tempdir()
fit_bb(counts_f = counts_f, N=10,
out = out, mc.cores=1)
unlink(out)

## End(Not run)

---

call gt helper, get posterior mean, expected gt and sd expected gt across all samples

Description

call gt helper, get posterior mean, expected gt and sd expected gt across all samples

Usage

gt_help(stan_samples, pop, data)

Arguments

stan_samples	matrix with samples extracted from stan fit object, params mu and lambda
pop	population to select AF for GT prior, defaults to EUR
data	data table 1 row with counts and EAF to apply model

Value

gt_help()

---

Pool randomly selected reads from different files

Description

Pool randomly selected reads from different files

Usage

poolreads(files, N = 1000, d = 10, out)

Arguments

files	names for files to extract reads
N	number of reads to extract
d	depth for reads
out	file name to save reads

Value

save files

Examples

counts_f <- system.file("extdata/input", "NA12878.chr22.Q20.allelicCounts.txt",
package = "bbmix",
mustWork = TRUE)

## In this example we only use one file and we take a pool of 10 reads

out <- tempfile()

poolreads(files=counts_f,
N=10,
d=10,
out = out)

unlink(out)

---