Mask sequences by thresholding the PWM

Mask sequences by thresholding the PWM. Sequences with a PWM above the threshold will be masked by 'N'. Sequences at the edges of the sequences will also be masked by 'N'.

Usage

mask_sequences_by_pwm(
  sequences,
  pssm,
  mask_thresh,
  pos_bits_thresh = 0.2,
  spat = NULL,
  spat_min = 0,
  spat_max = NULL,
  bidirect = TRUE,
  prior = 0.01
)

Arguments

sequences

a vector of sequences

pssm

a PSSM matrix or data frame. The columns of the matrix or data frame should be named with the nucleotides ('A', 'C', 'G' and 'T').

mask_thresh

Threshold for masking. Sequences with a PWM above this threshold will be masked by 'N'.

pos_bits_thresh

Mask only positions with amount of information contributed (Shannon entropy, measured in bits) above this threshold. The scale is the same as the y axis in the pssm logo plots.

spat

a data frame with the spatial model (as returned from the $spat slot from the regression). Should contain a column called 'bin' and a column called 'spat_factor'.

spat_min

the minimum position to use from the sequences. The default is 1.

spat_max

the maximum position to use from the sequences. The default is the length of the sequences.

bidirect

is the motif bi-directional. If TRUE, the reverse-complement of the motif will be used as well.

prior

a prior probability for each nucleotide.

Value

A vector with the masked sequences.

Examples

if (FALSE) { # \dontrun{
res <- regress_pwm(cluster_sequences_example, cluster_mat_example[, 1])
new_sequences <- mask_sequences_by_pwm(
    cluster_sequences_example,
    res$pssm,
    quantile(res$pred, 0.95),
    spat = res$spat
)

head(new_sequences)
} # }