reclanc

reclanc is a revival of ClaNC (Classification of microarrays to nearest centroids), by Alan R. Dabney.

Since the source has been lost (at least to my knowledge), the code comes from here with heavy modification.

reclanc is a nearest-centroid classifier for expression data. It tends to be a little more sensitive and accurate than similar models like PAM.

Besides its mere existence, reclanc differs slightly from the original ClaNC package in a few ways:

• reclanc supports a wide variety of inputs (data.frame, matrix, formula, recipe, ExpressionSet, and SummarizedExperiment)
• reclanc plays nicely with tidymodels, and offloads things like making folds to rsample and tuning to tune (see this vignette for how to leverage tidymodels with reclanc).
• Provides a prediction method based on correlation, rather than distance - useful for predicting classes from data from different sequencing modalities

Installation

You can install the development version of reclanc like so:

# install.packages("pak")
pak::pak("KaiAragaki/reclanc")

How to use it

library(reclanc)
lapply(synthetic_expression, head) # dummy data
#> $expression
#>        sample1  sample2  sample3  sample4  sample5  sample6  sample7  sample8
#> gene1 8.097529 7.119188 7.304400 7.554689 7.953206 7.714925 7.512700 8.597547
#> gene2 8.641837 9.400416 8.500865 8.878687 8.318438 8.728683 7.812591 7.638167
#> gene3 3.436236 4.317915 3.435193 3.515755 3.024976 4.762209 5.048956 2.006646
#> gene4 4.368008 5.212750 4.618249 4.201365 3.195294 4.707750 5.126769 6.178658
#> gene5 2.423974 3.563816 4.062362 2.163278 2.021435 2.813873 0.000000 4.652358
#> gene6 5.371205 5.919809 4.366915 4.805534 4.834856 5.622157 3.883531 3.593082
#>        sample9 sample10 sample11 sample12
#> gene1 6.475641 7.648858 8.637526 7.345038
#> gene2 8.110285 7.906104 7.424728 7.927039
#> gene3 2.739211 3.111668 3.161077 4.306611
#> gene4 5.170265 4.259578 5.872855 6.159023
#> gene5 1.532242 3.399823 3.691250 1.932937
#> gene6 4.246205 4.637316 3.575837 2.730452
#> 
#> $classes
#> [1] A A A A A A
#> Levels: A B

centroids <- clanc(
  synthetic_expression$expression,
  classes = synthetic_expression$classes,
  active = 5
)
centroids
#> <clanc> 
#> $centroids
#>    class   gene expression pooled_sd active prior
#> 1      A gene13   8.936462 0.3418472      5   0.5
#> 2      A gene21   7.379940 0.5279636      5   0.5
#> 3      A  gene2   8.744821 0.3147537      5   0.5
#> 4      A gene74   4.028507 0.4940783      5   0.5
#> 5      A gene41   4.328516 0.6317005      5   0.5
#> 6      A gene66   6.124761 0.5883218      5   0.5
#> 7      A gene24   4.307301 0.7214700      5   0.5
#> 8      A gene95   6.288173 0.4462475      5   0.5
#> 9      A gene94   7.777318 0.5375914      5   0.5
#> 10     A gene52   3.743798 0.5173769      5   0.5
#> 11     B gene13   9.938137 0.3418472      5   0.5
#> 12     B  gene2   8.273987 0.3147537      5   0.5
#> 13     B gene21   6.584681 0.5279636      5   0.5
#> 14     B gene41   5.518354 0.6317005      5   0.5
#> 15     B gene74   3.226598 0.4940783      5   0.5
#> 16     B gene24   3.370467 0.7214700      5   0.5
#> 17     B gene66   7.008174 0.5883218      5   0.5
#> 18     B gene94   8.422255 0.5375914      5   0.5
#> 19     B gene95   5.703161 0.4462475      5   0.5
#> 20     B gene52   2.438579 0.5173769      5   0.5

For information on basic usage, see this vignette. For a case study, as well as how to optimize the active parameter, see this vignette.

How it works

You can find a gentle introduction to how reclanc works here and a more in-depth and statistically rigorous definition of how this algorithm works in the original paper.

Citation

Citation for the original ClaNC paper:

Alan R. Dabney, Classification of microarrays to nearest centroids, Bioinformatics, Volume 21, Issue 22, November 2005, Pages 4148–4154, https://doi.org/10.1093/bioinformatics/bti681