Hi, I'd like to use curatedmetagenomicdata to download some experiments from ExperimentHub and am running into certificate error issues. I have previously searched this forum and tried the solutions I found, but they do not seem to work, as detailed in the following.  Immediately prior to R session, I tried deleting the corrupt cache from the command line and restarting R. I tried unlinking the cache during R, and I have verified that my bioconductor packages are current and valid.

The piece of code I am trying to execute successfully is:

library(ExperimentHub)

eh = ExperimentHub()

The errors and session info are detailed below. Any assistance is appreciated.

#Unix command prompt

$ rm -rf /Users/morxo28p/.ExperimentHub/

#From within R:

> file.exists('/Users/morxo28p/.ExperimentHub/experimenthub.sqlite3')

[1] FALSE

> library(ExperimentHub)

#library loaded successfully, terminal output of this trimmed

> eh = ExperimentHub()

updating metadata: retrieving 1 resource

Error: database is corrupt; remove it and try again

  database: ‘/Users/morxo28p//.ExperimentHub/experimenthub.sqlite3’

  reason: missing tables

In addition: Warning message:

download failed

  hub path: ‘https://experimenthub.bioconductor.org/metadata/experimenthub.sqlite3’

  cache path: ‘/Users/morxo28p//.ExperimentHub/experimenthub.sqlite3’

  reason: Peer certificate cannot be authenticated with given CA certificates 

> traceback()

9: stop("database is corrupt; remove it and try again", "\n  database: ", 

       sQuote(path), "\n  reason: ", conditionMessage(err), call. = FALSE)

8: value[[3L]](cond)

7: tryCatchOne(expr, names, parentenv, handlers[[1L]])

6: tryCatchList(expr, classes, parentenv, handlers)

5: tryCatch({

       expected <- c("biocversions", "input_sources", "location_prefixes", 

           "rdatapaths", "recipes", "resources", "statuses", "tags", 

           "timestamp")

       if (!all(expected %in% dbListTables(conn))) 

           stop("missing tables")

       sql <- "SELECT COUNT(id) FROM resources"

       if (.db_query(conn, sql)[[1]] == 0L) 

           warning("empty 'resources' table; database may be corrupt")

   }, error = function(err) {

       stop("database is corrupt; remove it and try again", "\n  database: ", 

           sQuote(path), "\n  reason: ", conditionMessage(err), 

           call. = FALSE)

   })

4: .db_is_valid(db_path)

3: .db_get(db_path, url, proxy)

2: .Hub("ExperimentHub", hub, cache, proxy, ...)

1: ExperimentHub()

#Troubleshooting - unlink and try again

> unlink('/Users/morxo28p/.ExperimentHub/experimenthub.sqlite3')

> file.exists('/Users/morxo28p/.ExperimentHub/experimenthub.sqlite3')

[1] FALSE

> eh = ExperimentHub()

updating metadata: retrieving 1 resource

Error: database is corrupt; remove it and try again

  database: ‘/Users/morxo28p//.ExperimentHub/experimenthub.sqlite3’

  reason: missing tables

In addition: Warning message:

download failed

  hub path: ‘https://experimenthub.bioconductor.org/metadata/experimenthub.sqlite3’

  cache path: ‘/Users/morxo28p//.ExperimentHub/experimenthub.sqlite3’

  reason: Peer certificate cannot be authenticated with given CA certificates 

# R session information

> BiocInstaller::biocValid()

[1] TRUE

> sessionInfo()

R version 3.3.2 (2016-10-31)

Platform: x86_64-apple-darwin13.4.0 (64-bit)

Running under: OS X El Capitan 10.11.6

locale:

[1] en_NZ.UTF-8/en_NZ.UTF-8/en_NZ.UTF-8/C/en_NZ.UTF-8/en_NZ.UTF-8

attached base packages:

[1] parallel  stats     graphics  grDevices utils     datasets  methods  

[8] base     

other attached packages:

[1] ExperimentHub_1.0.0 AnnotationHub_2.6.4 BiocGenerics_0.20.0

loaded via a namespace (and not attached):

 [1] Rcpp_0.12.8                   IRanges_2.8.1                

 [3] digest_0.6.10                 mime_0.5                     

 [5] R6_2.2.0                      xtable_1.8-2                 

 [7] DBI_0.5-1                     stats4_3.3.2                 

 [9] RSQLite_1.1-1                 BiocInstaller_1.24.0         

[11] httr_1.2.1                    curl_2.3                     

[13] S4Vectors_0.12.1              tools_3.3.2                  

[15] Biobase_2.34.0                shiny_0.14.2                 

[17] yaml_2.1.14                   httpuv_1.3.3                 

[19] AnnotationDbi_1.36.0          memoise_1.0.0                

[21] htmltools_0.3.5               interactiveDisplayBase_1.12.0

The certificate issue for 'https://experimenthub.bioconductor.org/' should now be resolved. Greg

I too see the issue when I call ExperimentHub – the CA certificates are related to Amazon S3, which is the storage mechanism on the backend. Although this should only trigger an warning and not a error. To fix the issue first do the following from the command line:

$ rm -rf /Users/morxo28p/.ExperimentHub

This should remove corrupt database and then you can do the following from R to download data from curatedMetagenomicData:

library(curatedMetagenomicData)
ZellerG_2014.metaphlan_bugs_list.stool()

It is also possible to bring up the names of datasets by typing:

curatedMetagenomicData::

And hitting tab for a list of autocompletion choices; although, that will only work in RStudio.

To be clear, this is not an issue related to curatedMetagenomicData, ExperimentHub, or AnnotationHub. CA certificates are an amazon problem and the corrupt database is a user issue. If further issues are encountered, try removing the database again and reinstalling packages.

UPDATE: Having followed my own advice, I too get the error and cannot fix the issue by the above steps. What I believe to be the issue is that Amazon S3 CA certificates have expired. If the download is absolutely essential at the moment you could cURL the sqlite file like this from command line:

$ mkdir /Users/morxo28p/.ExperimentHub
$ cd /Users/morxo28p/.ExperimentHub
$ curl -O -k  https://experimenthub.bioconductor.org/metadata/experimenthub.sqlite3

This will download the database, regardless of the expired CA certificates and then you should be able to download datasets from within R.

Lucas, it looks like the user has already followed your advice in their original post... I am having the same problem which seems to be an invalid certificate for https://experimenthub.bioconductor.org, an ExperimentHub issue. Val, do you know what the problem is?

Thank you Gregory, curatedMetagenomicData is working again for me!