Selection and Ranking of Omics Biomarkers Combinations Made Easy
Compute combinations.
Compute the combi score using glm(link='binomial') models and optional...
Reshape combiroc data in long format.
Show combinations with given markers.
Load CombiROC data.
Show distribution of intensity values for all the markers both singula...
Rank combinations.
Train logistic regression and compute ROC.
Load CombiROC data.
Show the composition of combinations.
Provide statistics for each marker.
Provides functions and a workflow to easily and powerfully calculating specificity, sensitivity and ROC curves of biomarkers combinations. Allows to rank and select multi-markers signatures as well as to find the best performing sub-signatures, now also from single-cell RNA-seq datasets. The method used was first published as a Shiny app and described in Mazzara et al. (2017) <doi:10.1038/srep45477> and further described in Bombaci & Rossi (2019) <doi:10.1007/978-1-4939-9164-8_16>, and widely expanded as a package as presented in the bioRxiv pre print Ferrari et al. <doi:10.1101/2022.01.17.476603>.
Useful links