Including continuous endpoints in Citrus Run

[SforR]

Dear all,

I have succesfully used the User Interface program and I get great results. However I am stuck with the next step. I want to include endpoints as continues values of the sampels but that is only possible when using the R script directly as mentioned in the "Getting started" protocol.

Does anyone has experience with that? I saw that there is a specific R script for the continuous endpoint ("citrus.continuousModel.R") but I don't know how to use this.

My first question is actually where and how to integrate the continues values. Am I right that this should be referring by the name "features" in the script? And should the format be similar as in the example CSV at the location: https://github.com/nolanlab/citrus/blob/master/inst/extdata/example2/citruskey.csv

I have no idea where to start. Can someone help me out with this? I am lacking some experience with R I suppose.

Best,

Sam

[SamGG]

Hi, I never used it, but I dived a little bit in the code. I think CSV of exemple5 is correct, other examples concern classification, ie building separated groups. Most of examples (1 to 4) concern classification, and family parameter is set by default to classification. I think you have to change it to continuous. You might look for the keyword continuous, and you will lead to the following functions and help messages written in R code. https://github.com/nolanlab/citrus/blob/master/R/citrus.full.R https://github.com/nolanlab/citrus/blob/master/R/citrus.model.R https://github.com/nolanlab/citrus/blob/master/R/citrus.continuousModel.R If you build an example script, that would be great. HTH

[SforR]

Hi Samuel,

Thank you for your input and thinking along. Example 5 is indeed what I am looking for, so that could be a hint. I have searched for example5 in the scripts but couldn't find anything. Now I will indeed search for "continuous".

Thanks again and I will keep you updated.

Sam

[SamGG]

By chance https://github.com/nolanlab/citrus/blob/master/inst/citrus.test.survival.R Best.

[SforR]

Samuel,

I played around with the R script based on the "runcitrus.R" script for the GUI output. Here I have changed the family to continuous and add different values in labels as.factor (see attachment for an example script).

When I run this script I see on the Y-ax the different values, however I don't know what the alghorithm exactly compares or calculates.

Best,

Sam

edited on 14-9-2018; as.factor(c("x1",x2"...) changed for as.numeric(c(n1,n2..)

Test.script.for.continuous.values.txt