dplyr::do error with surv_cutpoint

[MarcinKosinski]

dplyr::do error with surv_cutpoint

expected behaviour

surv_categorize works for grouped data.frames in dplyr::do worklfow.

actual behaviour

I get an error, probably from survival

# load neccessary packages
library(survminer)
library(survival)
library(RTCGA)
library(RTCGA.clinical)
library(RTCGA.mRNA)
# extract clinical information
clin <- survivalTCGA(BRCA.clinical, OV.clinical)
head(clin)

  times bcr_patient_barcode patient.vital_status
1  3767        TCGA-3C-AAAU                    0
2  3801        TCGA-3C-AALI                    0
3  1228        TCGA-3C-AALJ                    0
4  1217        TCGA-3C-AALK                    0
5   158        TCGA-4H-AAAK                    0
6  1477        TCGA-5L-AAT0                    0

# extract micro-array gene expression
expr <- expressionsTCGA(BRCA.mRNA, OV.mRNA, 
                        extract.cols = c("PAX8"))
head(expr)

# A tibble: 6 × 3
           bcr_patient_barcode   dataset     PAX8
                         <chr>     <chr>    <dbl>
1 TCGA-A1-A0SD-01A-11R-A115-07 BRCA.mRNA -0.54225
2 TCGA-A1-A0SE-01A-11R-A084-07 BRCA.mRNA -0.59475
3 TCGA-A1-A0SH-01A-11R-A084-07 BRCA.mRNA  0.49975
4 TCGA-A1-A0SJ-01A-11R-A084-07 BRCA.mRNA -0.58850
5 TCGA-A1-A0SK-01A-12R-A084-07 BRCA.mRNA -0.96475
6 TCGA-A1-A0SM-01A-11R-A084-07 BRCA.mRNA  0.57275

# join both datasets
library(tidyverse)
clin %>%
   left_join(expr %>%
       filter(substr(bcr_patient_barcode, 14, 15) == "01") %>% # tumour samples
       mutate(bcr_patient_barcode = substr(bcr_patient_barcode,1,12))) %>%
   filter(!is.na(PAX8)) %>% # no mRNA info
   mutate(dataset = gsub("\\..+", "", dataset)) -> # remove mRNA part from the name
   clin_expr
head(clin_expr)

  times bcr_patient_barcode patient.vital_status dataset     PAX8
1   437        TCGA-A1-A0SD                    0    BRCA -0.54225
2  1321        TCGA-A1-A0SE                    0    BRCA -0.59475
3  1437        TCGA-A1-A0SH                    0    BRCA  0.49975
4   416        TCGA-A1-A0SJ                    0    BRCA -0.58850
5   967        TCGA-A1-A0SK                    1    BRCA -0.96475
6   242        TCGA-A1-A0SM                    0    BRCA  0.57275

# perform surv_categorize in groups determined by dataset
clin_expr %>%
   group_by(dataset) %>%
   do(data_cat = 
         surv_cutpoint(data = ., time = "times",
              event = "patient.vital_status",
              variables = c("PAX8", "dataset"))) ->
   clin_exp_categorized

Error in survival::Surv(time, event): Time variable is not numeric

while this works for not grouped set

surv_cutpoint(data = clin_expr, time = "times",
              event = "patient.vital_status",
              variables = c("PAX8", "dataset")) ->
   clin_exp_categorized
head(surv_categorize(clin_exp_categorized))

  times patient.vital_status PAX8 dataset
1   437                    0  low    BRCA
2  1321                    0  low    BRCA
3  1437                    0  low    BRCA
4   416                    0  low    BRCA
5   967                    1  low    BRCA
6   242                    0  low    BRCA

session info

sessionInfo()

R version 3.3.2 (2016-10-31)
Platform: x86_64-pc-linux-gnu (64-bit)
Running under: Ubuntu 14.04.5 LTS

locale:
 [1] LC_CTYPE=pl_PL.UTF-8       LC_NUMERIC=C               LC_TIME=pl_PL.UTF-8        LC_COLLATE=pl_PL.UTF-8     LC_MONETARY=pl_PL.UTF-8   
 [6] LC_MESSAGES=pl_PL.UTF-8    LC_PAPER=pl_PL.UTF-8       LC_NAME=C                  LC_ADDRESS=C               LC_TELEPHONE=C            
[11] LC_MEASUREMENT=pl_PL.UTF-8 LC_IDENTIFICATION=C       

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base     

other attached packages:
 [1] dplyr_0.5.0                 purrr_0.2.2                 readr_1.0.0                 tidyr_0.6.0                 tibble_1.2                 
 [6] tidyverse_1.0.0             RTCGA.mRNA_1.2.0            RTCGA.clinical_20151101.4.0 RTCGA_1.5.1                 survival_2.40-1            
[11] survminer_0.2.4             ggplot2_2.2.1               knitr_1.15.1               

loaded via a namespace (and not attached):
 [1] Rcpp_0.12.8.4         plyr_1.8.4            viridis_0.3.4         tools_3.3.2           digest_0.6.11         evaluate_0.10        
 [7] gtable_0.2.0          lattice_0.20-34       Matrix_1.2-7.1        DBI_0.5-1             yaml_2.1.14           mvtnorm_1.0-5        
[13] gridExtra_2.2.1       stringr_1.1.0         httr_1.2.1            xml2_1.0.0.9002       exactRankTests_0.8-28 maxstat_0.7-24       
[19] rprojroot_1.1         grid_3.3.2            data.table_1.10.0     R6_2.2.0              XML_3.98-1.5          rmarkdown_1.3        
[25] magrittr_1.5          backports_1.0.4       scales_0.4.1          htmltools_0.3.5       ggthemes_3.3.0        splines_3.3.2        
[31] assertthat_0.1        rvest_0.3.2           colorspace_1.3-2      stringi_1.1.2         lazyeval_0.2.0        munsell_0.4.3

[kassambara]

This issue is now fixed. The R package maxstat doesn't support very well an object of class tbl_df. Now, in surv_cutpoint() the input data is systematically transformed into a standard data.frame format

# perform surv_categorize in groups determined by dataset
clin_expr %>%
   group_by(dataset) %>%
   do(data_cat = 
         surv_cutpoint(data = ., time = "times",
              event = "patient.vital_status",
              variables = c("PAX8", "dataset"))) ->
   clin_exp_categorized

[MarcinKosinski]

Thank you @kassambara for your kind help.
You can close the issue :)