fb_diploid: Inferring hidden ancestry states from diploids
In tylerhether/fbgenotyper: R package using the Forward-Backward algorithm to infer genotypes, recombination hotspots, and gene conversion tracts from low-coverage next-generation sequence data

Description Usage Arguments Details Value Author(s) References See Also Examples

Uses the forward-backward algorithm to estimate ancestry along a given chromosome for a given genotyped diploid.

1	fb_diploid(snp_locations, p0, p1, p_assign, scale)

`snp_locations`	a numeric vector specifying the locations of each snp (in bps). This vector is assumed to be ordered (sorted from smallest to largest snp).
`p0`	a vector specifying the number of reads that mapped to parent 0. p0 is assumed to be in the smae order as snp_locations.
`p1`	a vector specifying the number of reads that mapped to parent 1. p1 is assumed to be in the smae order as snp_locations.
`p_assign`	a value specifying the assignment probabilty (see details).
`scale`	a numeric specifying the the genome-wide recombination rate (Morgans / bp). `scale` is assumed to be between 0 and 1 but in practice it is usually quite small.

This is an extension of fb_haploid with three hidden states: two homozygous states and a heterozygous state.

a dataframe with the following columns:

snp_locations
p0
p1
The emission matrix (3 columns) specifying the emmision probability of belong to the only parent 0 (emiss.1), both parents (emiss.1), and only parent 2 (emiss.3)
the forward probability matrix (3 columns) giving the scaled forward probabilities of belong to the only parent 0 (forward.1), both parents (forward.2), and only parent 2 (forward.3)
the backward probability matrix (3 columns) giving the scaled backward probabilities of belong to the only parent 0 (backward.1), both parents (backward.2), and only parent 2 (backward.3)
Fscale The forward scaling factor for each snp
Bscale The backward scaling factor for each snp
posterior probability matrix (3 columns) posterior probabilities for belong to each parental state
states_inferred a vector of length snp_locations giving the state with the highest posterior.
lnL a numeric giving the total likelihood of the data (see fb_haploid ).

Tyler D. Hether

Hether, T.D., C. G. Wiench1, and P.A. Hohenlohe (in review). 2015. Novel molecular and analytical tools for efficient estimation of rates of meiotic crossover, non-crossover and gene conversion

Drubin, R. S. Eddy, A. Krogh, and G. Mitchison. 1998. Biological Sequence Analysis: Probabilistic Models of proteins and nucleic acids. Cambridge University Press, Cambridge CB2 8RU, UK.

fb_haploid

set.seed(1234567)        # For reproducibility
n_spores <- 1            # number of spores
l <- 75                  # number of snps to simulate
c <- 3.5e-05             # recombination rate between snps (Morgan/bp)
snps <- c(1:l)*1.3e4     # snps are evenly spaced 20kbp apart
p_a <- 0.95              # assignment probability
coverage <- 2.1          # mean coverage
# Now simulate two haploids
sim1 <- sim_en_masse(n.spores=n_spores, scale=c, snps=snps,
 p.assign=p_a, mu.rate=0, f.cross=0.8, f.convert=0.3,
 length.conversion=2e3, coverage=coverage)
sim2 <- sim_en_masse(n.spores=n_spores, scale=c, snps=snps,
 p.assign=p_a, mu.rate=0, f.cross=0.8, f.convert=0.3,
 length.conversion=2e3, coverage=coverage)
# Now merge the two haploids to make a diploid
p0 <- sim1$p0+sim2$p0
p1 <- sim1$p1+sim2$p1
res <- fb_diploid(snp_locations=sim1$Snp, p0=p0, p1=p1, p_assign=p_a, scale=c)
res