col_mode

<p>Determining the most common character in each column.</p>

internal

The 'debar' sequence processing pipeline is designed for denoising high throughput
sequencing data for the animal DNA barcode marker cytochrome c oxidase I (COI). The package
is designed to detect and correct insertion and deletion errors within sequencer outputs.
This is accomplished through comparison of input sequences against a profile hidden Markov
model (PHMM) using the Viterbi algorithm (for algorithm details see Durbin et al. 1998,
ISBN: 9780521629713). Inserted base pairs are removed and deleted base pairs are accounted
for through the introduction of a placeholder character. Since the PHMM is a probabilistic
representation of the COI barcode, corrections are not always perfect. For this reason
'debar' censors base pairs adjacent to reported indel sites, turning them into placeholder
characters (default is 7 base pairs in either direction, this feature can be disabled).
Testing has shown that this censorship results in the correct sequence length being restored,
and erroneous base pairs being masked the vast majority of the time (>95%).

Cameron Nugent

col_mode: Determining the most common character in each column.

Description

Usage

Arguments

Value