callAllele: Automatically call or flag alleles from NGS data

In romunov/fishbone: Tools for Calling Alleles/Genotypes from Illumina NGS Sequences

Description

Call or flag candidate alleles to construct a consensus genotype.

Usage

callAllele(fb, tbase = NULL, clean = TRUE, verbose = FALSE)

Arguments

fb	A fishbone object.
tbase	A data.frame with thresholds which are locus specific. Thresholds are: stutter (if lower than this, allele is ignored as stutter) disbalance (if heterozygous alleles are not in 1:1 ratio) low count (anything below this threshold gets flagged as light on read count) allele with no stutter height (if no stutter is found, how many reads do we allow for alleles to be called)
clean	Logical. If TRUE (default), it will return only called alleles and their stutters.
verbose	Logical. If TRUE, it will print which sample is being processed.

Details

Function works on an individual run (one PCR reaction). Apply this to sample * locus * run combination. The data should come from an NGS run as processed by de Barba et al. (2016).

For algorithm used to find stutters, see findStutter.

De Barba, M., Miquel, C., Lobréaux, S., Quenette, P. Y., Swenson, J. E., & Taberlet, P. (2016). High-throughput microsatellite genotyping in ecology: improved accuracy, efficiency, standardization and success with low-quantity and degraded DNA. Molecular Ecology Resources, 1-16. https://doi.org/10.1111/1755-0998.12594

Key for abbreviations used in (pseudo)code:

• A = allele

• S = stutter

• R = relative low threshold

• L = low count threshold

• D = disbalance

The result is appended three columns; one for called A, one for flagged alleles and if read is a stutter. Possible flags are:

• L = low amplification threshold (if for some reason, number of total reads is very low, alleles get a flag)

• N = no stutter (if there was enough reads but no stutter was present)

• D = disbalance - alleles not in balance (expecting 1:1 for heterozygotes, those out of balance flagged)

• M = multiple alleles (self explanatory)

Algorithm is as follows:

• 0. find max allele height

• 1. if allele has number of reads < L, flag it as "L"

• 2. see if allele has stutter

• 2a. if yes, mark as called

• 2aa. if A in disbalance (A < D), flag as "D"

• 2ab. mark stutter as such $stutter = TRUE

• 2b. if no, check AlleleWithNoStutterHeight

• 2ba. if x > AlleleWithNoStutterHeight, add flag "N"

• 2bb. if x < AlleleWithNoStutterHeight, ignore allele

• 3. if number of unflagged alleles is more than 2 (those marked with D are not counted), add flag "M" to all

Value

Output should be all alleles and their stutters.

romunov/fishbone documentation built on June 17, 2019, 2:06 a.m.