Data from the cBioPortal API (cBioPortalData())

Typically, the number of errors encountered via the API are low. There are only a handful of packages that error when we apply the utility functions to provide a MultiAssayExperiment data representation.

First, we load the error Rda dataset.

api_errs <- system.file(
    "extdata", "api", "err_api_info.rda",
    package = "cBioPortalData", mustWork = TRUE
)
load(api_errs)

We can now inspect the contents of the data:

class(err_api_info)

## [1] "list"

length(err_api_info)

## [1] 6

lengths(err_api_info)

##                           Barcodes must start with 'TCGA' 
##                                                         2 
##                     group length is 0 but data length > 0 
##                                                         1 
##   Frequency of NA values higher than the cutoff tolerance 
##                                                         2 
##                          Inconsistent build numbers found 
##                                                        33 
##         `n` must be a single number, not an integer `NA`. 
##                                                         1 
## Argument 1 must be a data frame or a named atomic vector. 
##                                                         1

There were about 6 unique errors during the last build run.

names(err_api_info)

## [1] "Barcodes must start with 'TCGA'"                          
## [2] "group length is 0 but data length > 0"                    
## [3] "Frequency of NA values higher than the cutoff tolerance"  
## [4] "Inconsistent build numbers found"                         
## [5] "`n` must be a single number, not an integer `NA`."        
## [6] "Argument 1 must be a data frame or a named atomic vector."

The most common error was Inconsistent build numbers found. This is due to annotations from different build numbers that were not able to be resolved.

To see what datasets (cancer_study_id s) have that error we can use:

err_api_info[['Inconsistent build numbers found']]

##  [1] "msk_ch_2020"                       "msk_access_2021"                  
##  [3] "mixed_msk_tcga_2021"               "mixed_impact_subset_2022"         
##  [5] "pan_origimed_2020"                 "prad_msk_stopsack_2021"           
##  [7] "pancan_pcawg_2020"                 "prad_pik3r1_msk_2021"             
##  [9] "skcm_tcga"                         "stad_tcga"                        
## [11] "stad_tcga_pub"                     "skcm_tcga_pan_can_atlas_2018"     
## [13] "stad_tcga_pan_can_atlas_2018"      "stes_tcga_pub"                    
## [15] "summit_2018"                       "cfdna_msk_2019"                   
## [17] "blca_bcan_hcrn_2022"               "nsclc_ctdx_msk_2022"              
## [19] "thyroid_mskcc_2016"                "skcm_mskcc_2014"                  
## [21] "tmb_mskcc_2018"                    "rectal_msk_2019"                  
## [23] "skcm_tcga_pub_2015"                "msk_spectrum_tme_2022"            
## [25] "ucec_ccr_cfdna_msk_2022"           "paired_bladder_2022"              
## [27] "mtnn_msk_2022"                     "pog570_bcgsc_2020"                
## [29] "sarcoma_msk_2023"                  "bowel_colitis_msk_2022"           
## [31] "luad_mskcc_2023_met_organotropism" "coad_silu_2022"                   
## [33] "paac_msk_jco_2023"

We can also have a look at the entirety of the dataset.

err_api_info

## $`Barcodes must start with 'TCGA'`
## [1] "blca_msk_tcga_2020"    "nsclc_tcga_broad_2016"
## 
## $`group length is 0 but data length > 0`
## [1] "glioma_msk_2018"
## 
## $`Frequency of NA values higher than the cutoff tolerance`
## [1] "mixed_selpercatinib_2020" "ucec_ccr_msk_2022"       
## 
## $`Inconsistent build numbers found`
##  [1] "msk_ch_2020"                       "msk_access_2021"                  
##  [3] "mixed_msk_tcga_2021"               "mixed_impact_subset_2022"         
##  [5] "pan_origimed_2020"                 "prad_msk_stopsack_2021"           
##  [7] "pancan_pcawg_2020"                 "prad_pik3r1_msk_2021"             
##  [9] "skcm_tcga"                         "stad_tcga"                        
## [11] "stad_tcga_pub"                     "skcm_tcga_pan_can_atlas_2018"     
## [13] "stad_tcga_pan_can_atlas_2018"      "stes_tcga_pub"                    
## [15] "summit_2018"                       "cfdna_msk_2019"                   
## [17] "blca_bcan_hcrn_2022"               "nsclc_ctdx_msk_2022"              
## [19] "thyroid_mskcc_2016"                "skcm_mskcc_2014"                  
## [21] "tmb_mskcc_2018"                    "rectal_msk_2019"                  
## [23] "skcm_tcga_pub_2015"                "msk_spectrum_tme_2022"            
## [25] "ucec_ccr_cfdna_msk_2022"           "paired_bladder_2022"              
## [27] "mtnn_msk_2022"                     "pog570_bcgsc_2020"                
## [29] "sarcoma_msk_2023"                  "bowel_colitis_msk_2022"           
## [31] "luad_mskcc_2023_met_organotropism" "coad_silu_2022"                   
## [33] "paac_msk_jco_2023"                
## 
## $``n` must be a single number, not an integer `NA`.`
## [1] "msk_met_2021"
## 
## $`Argument 1 must be a data frame or a named atomic vector.`
## [1] "makeanimpact_ccr_2023"

Packaged data from cBioDataPack()

Now let’s look at the errors in the packaged datasets that are used for cBioDataPack:

pack_errs <- system.file(
    "extdata", "pack", "err_pack_info.rda",
    package = "cBioPortalData", mustWork = TRUE
)
load(pack_errs)

We can do the same for this data:

length(err_pack_info)

## [1] 5

lengths(err_pack_info)

##                                         more columns than column names 
##                                                                      9 
##                Frequency of NA values higher than the cutoff tolerance 
##                                                                      5 
##                                                 non-character argument 
##                                                                      2 
## invalid class "ExperimentList" object: \n    Non-unique names provided 
##                                                                      2 
##                                    'wget' call had nonzero exit status 
##                                                                     11

We can get a list of all the errors present:

names(err_pack_info)

## [1] "more columns than column names"                                          
## [2] "Frequency of NA values higher than the cutoff tolerance"                 
## [3] "non-character argument"                                                  
## [4] "invalid class \"ExperimentList\" object: \n    Non-unique names provided"
## [5] "'wget' call had nonzero exit status"

And finally the full list of errors:

err_pack_info

## $`more columns than column names`
## [1] "ccrcc_utokyo_2013"                "coadread_tcga_pan_can_atlas_2018"
## [3] "gbm_cptac_2021"                   "ov_tcga_pan_can_atlas_2018"      
## [5] "pan_origimed_2020"                "sarc_tcga_pan_can_atlas_2018"    
## [7] "luad_mskimpact_2021"              "mbl_dkfz_2017"                   
## [9] "brca_tcga_pan_can_atlas_2018"    
## 
## $`Frequency of NA values higher than the cutoff tolerance`
## [1] "ihch_mskcc_2020"          "ihch_msk_2021"           
## [3] "mixed_selpercatinib_2020" "mixed_msk_tcga_2021"     
## [5] "ucec_ccr_msk_2022"       
## 
## $`non-character argument`
## [1] "mbn_mdacc_2013" "pcpg_tcga_pub" 
## 
## $`invalid class "ExperimentList" object: \n    Non-unique names provided`
## [1] "stad_tcga_pub" "mpnst_mskcc"  
## 
## $`'wget' call had nonzero exit status`
##  [1] "makeanimpact_ccr_2023"   "prad_organoids_msk_2022"
##  [3] "mtnn_msk_2022"           "sarcoma_msk_2023"       
##  [5] "bowel_colitis_msk_2022"  "bladder_mskcc_2022"     
##  [7] "paac_msk_jco_2023"       "nbl_msk_2023"           
##  [9] "rms_msk_2023"            "gist_msk_2023"          
## [11] "egc_trap_ccr_msk_2023"