vignettes/selex-inference.md

In anthony-aylward/selex: Multinomial Logistic Regrestion for SELEX sequencing data

title: "Statistical inference on SELEX sequencing data" author: "Anthony Aylward" date: "2018-07-28" output: rmarkdown::html_vignette vignette: > %\VignetteIndexEntry{Vignette Title} %\VignetteEngine{knitr::rmarkdown} %\VignetteEncoding{UTF-8}

The selex package includes an example of SELEX count data.

library(selex)
example_counts
#>    a  c  g  t
#> 1 25 25 25 25
#> 2 50 20 15 15
#> 3 75  8  7  7
#> 4 95  2  1  1
#> 5 95  2  1  1

First, fit a multinomial logit regression model to the counts.

fit <- selex_multinom(example_counts, weights = c(8, 1, 2, 4, 8), ref = "c")
summary(fit)
#> Call:
#> multinom(formula = counts ~ cycle, weights = weights)
#> 
#> Coefficients:
#>   (Intercept)      cycle
#> a 0.011978665  1.0665507
#> g 0.002133017 -0.1765759
#> t 0.002638545 -0.1766294
#> 
#> Std. Errors:
#>   (Intercept)      cycle
#> a  0.09531140 0.06128190
#> g  0.09787694 0.09098062
#> t  0.09786460 0.09097057
#> 
#> Residual Deviance: 3288.078 
#> AIC: 3300.078

Then, numerically compute p-values for the coefficients.

selex_pvals(fit)
#>         a         g         t 
#> 0.0000000 0.6271237 0.5765662

anthony-aylward/selex documentation built on May 21, 2019, 1:42 a.m.