Separate 2 groups in Cox regression

Instalation

if (!require("BiocManager")) {
    install.packages("BiocManager")
}
BiocManager::install("glmSparseNet")

Required Packages

library(futile.logger)
library(ggplot2)
library(glmSparseNet)
library(survival)

# Some general options for futile.logger the debugging package
flog.layout(layout.format("[~l] ~m"))
options("glmSparseNet.show_message" = FALSE)
# Setting ggplot2 default theme as minimal
theme_set(ggplot2::theme_minimal())

Prepare data

data("cancer", package = "survival")
xdata <- survival::ovarian[, c("age", "resid.ds")]
ydata <- data.frame(
    time = survival::ovarian$futime,
    status = survival::ovarian$fustat
)

Separate using age as co-variate

(group cutoff is median calculated relative risk)

resAge <- separate2GroupsCox(c(age = 1, 0), xdata, ydata)

Kaplan-Meier survival results

## Call: survfit(formula = survival::Surv(time, status) ~ group, data = prognosticIndexDf)
## 
##                n events median 0.95LCL 0.95UCL
## Low risk - 1  13      4     NA     638      NA
## High risk - 1 13      8    464     268      NA

Plot

A individual is attributed to low-risk group if its calculated relative risk (using Cox Proportional model) is below or equal the median risk.

The opposite for the high-risk groups, populated with individuals above the median relative-risk.

Separate using age as co-variate (group cutoff is 40% - 60%)

resAge4060 <-
    separate2GroupsCox(c(age = 1, 0),
        xdata,
        ydata,
        probs = c(.4, .6)
    )

Kaplan-Meier survival results

## Call: survfit(formula = survival::Surv(time, status) ~ group, data = prognosticIndexDf)
## 
##                n events median 0.95LCL 0.95UCL
## Low risk - 1  11      3     NA     563      NA
## High risk - 1 10      7    359     156      NA

Plot

A individual is attributed to low-risk group if its calculated relative risk (using Cox Proportional model) is below the median risk.

The opposite for the high-risk groups, populated with individuals above the median relative-risk.

Separate using age as co-variate (group cutoff is 60% - 40%)

This is a special case where you want to use a cutoff that includes some sample on both high and low risks groups.

resAge6040 <- separate2GroupsCox(
    chosenBetas = c(age = 1, 0),
    xdata,
    ydata,
    probs = c(.6, .4),
    stopWhenOverlap = FALSE
)
## Warning in buildPrognosticIndexDataFrame(ydata, probs, stopWhenOverlap, : The cutoff values given to the function allow for some over samples in both groups, with:
##   high risk size (15) + low risk size (16) not equal to xdata/ydata rows (31 != 26)
## 
## We are continuing with execution as parameter `stopWhenOverlap` is FALSE.
##   note: This adds duplicate samples to ydata and xdata xdata

Kaplan-Meier survival results

## Kaplan-Meier results
## Call: survfit(formula = survival::Surv(time, status) ~ group, data = prognosticIndexDf)
## 
##                n events median 0.95LCL 0.95UCL
## Low risk - 1  16      5     NA     638      NA
## High risk - 1 15      9    475     353      NA

Plot

A individual is attributed to low-risk group if its calculated relative risk (using Cox Proportional model) is below the median risk.

The opposite for the high-risk groups, populated with individuals above the median relative-risk.

Session Info

sessionInfo()
## R version 4.4.2 (2024-10-31)
## Platform: x86_64-pc-linux-gnu
## Running under: Ubuntu 24.04.1 LTS
## 
## Matrix products: default
## BLAS:   /usr/lib/x86_64-linux-gnu/openblas-pthread/libblas.so.3 
## LAPACK: /usr/lib/x86_64-linux-gnu/openblas-pthread/libopenblasp-r0.3.26.so;  LAPACK version 3.12.0
## 
## locale:
##  [1] LC_CTYPE=en_US.UTF-8       LC_NUMERIC=C              
##  [3] LC_TIME=en_US.UTF-8        LC_COLLATE=C              
##  [5] LC_MONETARY=en_US.UTF-8    LC_MESSAGES=en_US.UTF-8   
##  [7] LC_PAPER=en_US.UTF-8       LC_NAME=C                 
##  [9] LC_ADDRESS=C               LC_TELEPHONE=C            
## [11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C       
## 
## time zone: Etc/UTC
## tzcode source: system (glibc)
## 
## attached base packages:
##  [1] grid      parallel  stats4    stats     graphics  grDevices utils    
##  [8] datasets  methods   base     
## 
## other attached packages:
##  [1] glmnet_4.1-8                VennDiagram_1.7.3          
##  [3] reshape2_1.4.4              forcats_1.0.0              
##  [5] Matrix_1.7-1                glmSparseNet_1.25.0        
##  [7] TCGAutils_1.27.6            curatedTCGAData_1.28.1     
##  [9] MultiAssayExperiment_1.33.4 SummarizedExperiment_1.37.0
## [11] Biobase_2.67.0              GenomicRanges_1.59.1       
## [13] GenomeInfoDb_1.43.2         IRanges_2.41.2             
## [15] S4Vectors_0.45.2            BiocGenerics_0.53.3        
## [17] generics_0.1.3              MatrixGenerics_1.19.0      
## [19] matrixStats_1.4.1           futile.logger_1.4.3        
## [21] survival_3.8-3              ggplot2_3.5.1              
## [23] dplyr_1.1.4                 BiocStyle_2.35.0           
## 
## loaded via a namespace (and not attached):
##   [1] sys_3.4.3                 jsonlite_1.8.9           
##   [3] shape_1.4.6.1             magrittr_2.0.3           
##   [5] GenomicFeatures_1.59.1    farver_2.1.2             
##   [7] rmarkdown_2.29            BiocIO_1.17.1            
##   [9] zlibbioc_1.52.0           vctrs_0.6.5              
##  [11] memoise_2.0.1             Rsamtools_2.23.1         
##  [13] RCurl_1.98-1.16           rstatix_0.7.2            
##  [15] htmltools_0.5.8.1         S4Arrays_1.7.1           
##  [17] BiocBaseUtils_1.9.0       progress_1.2.3           
##  [19] AnnotationHub_3.15.0      lambda.r_1.2.4           
##  [21] curl_6.0.1                broom_1.0.7              
##  [23] Formula_1.2-5             SparseArray_1.7.2        
##  [25] pROC_1.18.5               sass_0.4.9               
##  [27] bslib_0.8.0               plyr_1.8.9               
##  [29] httr2_1.0.7               zoo_1.8-12               
##  [31] futile.options_1.0.1      cachem_1.1.0             
##  [33] buildtools_1.0.0          GenomicAlignments_1.43.0 
##  [35] mime_0.12                 lifecycle_1.0.4          
##  [37] iterators_1.0.14          pkgconfig_2.0.3          
##  [39] R6_2.5.1                  fastmap_1.2.0            
##  [41] GenomeInfoDbData_1.2.13   digest_0.6.37            
##  [43] colorspace_2.1-1          AnnotationDbi_1.69.0     
##  [45] ExperimentHub_2.15.0      RSQLite_2.3.9            
##  [47] ggpubr_0.6.0              filelock_1.0.3           
##  [49] labeling_0.4.3            km.ci_0.5-6              
##  [51] httr_1.4.7                abind_1.4-8              
##  [53] compiler_4.4.2            bit64_4.5.2              
##  [55] withr_3.0.2               backports_1.5.0          
##  [57] BiocParallel_1.41.0       carData_3.0-5            
##  [59] DBI_1.2.3                 ggsignif_0.6.4           
##  [61] biomaRt_2.63.0            rappdirs_0.3.3           
##  [63] DelayedArray_0.33.3       rjson_0.2.23             
##  [65] tools_4.4.2               glue_1.8.0               
##  [67] restfulr_0.0.15           checkmate_2.3.2          
##  [69] gtable_0.3.6              KMsurv_0.1-5             
##  [71] tzdb_0.4.0                tidyr_1.3.1              
##  [73] survminer_0.5.0           data.table_1.16.4        
##  [75] hms_1.1.3                 car_3.1-3                
##  [77] xml2_1.3.6                XVector_0.47.1           
##  [79] BiocVersion_3.21.1        foreach_1.5.2            
##  [81] pillar_1.10.0             stringr_1.5.1            
##  [83] splines_4.4.2             BiocFileCache_2.15.0     
##  [85] lattice_0.22-6            rtracklayer_1.67.0       
##  [87] bit_4.5.0.1               tidyselect_1.2.1         
##  [89] maketools_1.3.1           Biostrings_2.75.3        
##  [91] knitr_1.49                gridExtra_2.3            
##  [93] xfun_0.49                 stringi_1.8.4            
##  [95] UCSC.utils_1.3.0          yaml_2.3.10              
##  [97] evaluate_1.0.1            codetools_0.2-20         
##  [99] tibble_3.2.1              BiocManager_1.30.25      
## [101] cli_3.6.3                 xtable_1.8-4             
## [103] munsell_0.5.1             jquerylib_0.1.4          
## [105] survMisc_0.5.6            Rcpp_1.0.13-1            
## [107] GenomicDataCommons_1.31.0 dbplyr_2.5.0             
## [109] png_0.1-8                 XML_3.99-0.17            
## [111] readr_2.1.5               blob_1.2.4               
## [113] prettyunits_1.2.0         bitops_1.0-9             
## [115] scales_1.3.0              purrr_1.0.2              
## [117] crayon_1.5.3              rlang_1.1.4              
## [119] KEGGREST_1.47.0           rvest_1.0.4              
## [121] formatR_1.14