View source: R/signatureIdentificationLib.R
fitSignaturesWithSignatures | R Documentation |
This function attempts to reconstruct a set of signatures (signatures_to_fit) using a second set of signatures (signatures_to use), either just checking the similarity between the signatures of the first and the second set, or using a linear combination of signatures in the second set to model the first set. The typical use of this function would be to check whether a set of extracted signatures can be matched to a set of known signatures, a procedure often called signature assignment or signature identification.
fitSignaturesWithSignatures(
signatures_to_fit,
signatures_to_use,
max_combinations = 2,
sig_min_contrib = 0.15,
minsim = 0.75,
max_from_each_class = 10,
nparallel = 1
)
signatures_to_fit |
matrix with the mutational signatures that need to be identified/fitted. Each column has a signature and each row corresponds to a mutational channel. Columns need to sum to 1. |
signatures_to_use |
matrix of mutational signatures to be used as a reference and that will be fitted/combined to match the signatures_to_fit. Each column has a signature and each row corresponds to a mutational channel. Columns need to sum to 1. |
max_combinations |
maximum number of signatures_to_use to be used as a linear combination to fit each signatures_to_fit |
sig_min_contrib |
minimum percentage of contribution that a signatures_to_use should have to a linear combination when matching a signature in signatures_to_fit |
minsim |
minimum cosine similarity between a signatures_to_fit and a match, where a match is either a signature in signatures_to_use or a linear combination of signatures in signatures_to_use |
max_from_each_class |
maximum number of matches to consider in each class, where the class indicates the number of signatures in a linear combination. If there are more matches in the class than the maximum specified, then the matches with the highest cosine similarity will be reported |
nparallel |
how many parallel cpus to use |
tables with the best linear combinations of signatures in signatures_to_use that match signatures in signatures_to_fit
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.