Error when importing EPIC data to RnBeads

262 views
Skip to first unread message

6279...@qq.com

unread,
Jul 17, 2018, 10:32:29 PM7/17/18
to Epigenomics forum
Hello there,

I met two problems when importing EPIC beta value to RnBeads.
(1) Problem 1
Here is my code :
data.dir <- file.path(tumor.path,"RnBeads/data")
data
.source1 <- c(file.path(data.dir, "Sinfo of Tumor Normal for RnBeads.csv"),file.path(data.dir,"1493_Malouf-beta_values-noob for RnBeads.csv"))
data
.source2 <- c(file.path(data.dir, "Sinfo of Tumor C1 C2 for RnBeads.csv"),file.path(data.dir,"1493_Malouf-beta_values-noob for RnBeads.csv"))
# Directory where the output should be written to
analysis
.dir <- file.path(tumor.path,"RnBeads/analysis")
# Directory where the report files should be written to
report
.dir <- file.path(analysis.dir, "reports")
result
<- rnb.run.import(data.source=data.source1, data.type="infinium.data.files", dir.reports=report.dir


And I got:
2018-07-18 10:13:54     8.1  STATUS                             STARTED Loading Data
2018-07-18 10:13:54     8.1    INFO                                 Number of cores: 1
2018-07-18 10:13:54     8.1    INFO                                 Loading data of type "infinium.data.files"
2018-07-18 10:13:54     8.1  STATUS                                 STARTED Loading Data from Table Files
2018-07-18 10:13:54     8.1    INFO                                     Read table with sample information
2018-07-18 10:14:51     9.3    INFO                                     Read table with betas
Error in match.probes2annotation(probes, target, assembly) : 
  probes contains invalid IDs

I think RnBeads didn't recognize the EPIC data and consider it as a 450k data, but I didn't find a parameter in rnb.run.import to choose annotation type.
(2)Problem 2:
I am really appreciated that you made a GUI tool and I can choose EPIC data type, it seems the GUI tool can only import IDAT file, and when I use IDAT file, I got another error:
> rnb.run.dj()

Listening on http://127.0.0.1:4478
2018-07-18 10:00:28     8.0  STATUS STARTED RnBeads Pipeline
2018-07-18 10:00:28     8.0    INFO     Analysis Title: RnBeads Analysis
2018-07-18 10:00:28     8.0    INFO     Initialized report index and saved to index.html
2018-07-18 10:00:28     8.0  STATUS     STARTED Loading Data
2018-07-18 10:00:28     8.0    INFO         Number of cores: 1
2018-07-18 10:00:29     8.0    INFO         Loading data of type "idat.dir"
2018-07-18 10:00:29     8.0  STATUS         STARTED Loading Data from IDAT Files
2018-07-18 10:00:29     8.0   ERROR             sample.sheet does not contain column barcode
Warning: Error in logger.error: sample.sheet does not contain column barcode

I got the same error when I use R cmd to import EPIC IDAT file:
2018-07-18 10:12:45     8.1  STATUS                     STARTED Loading Data
2018-07-18 10:12:45     8.1    INFO                         Number of cores: 1
2018-07-18 10:12:45     8.1    INFO                         Loading data of type "infinium.idat.dir"
2018-07-18 10:12:45     8.1  STATUS                         STARTED Loading Data from IDAT Files
2018-07-18 10:12:45     8.1   ERROR                             sample.sheet does not contain column barcode
Error in logger.error(txt) : sample.sheet does not contain column barcode

Now I don't know how to fix it ...
I would be greatly appreciated if someone could help me.
I

Michael Scherer

unread,
Jul 18, 2018, 3:30:02 AM7/18/18
to Epigenomics forum
Hey,

To detect the reason for your first problem, I would need to know what kind of data you'd like to import into RnBeads. Is it preprocessed beta-values in the form of the tabular file (e.g. CSV)? If yes, what is the row count of this matrix and does the matrix contain row and column names? If you have the IDAT files available, I would anyway start with those and leave preprocessing and quality control to RnBeads.

For your second problem: For Infinium data, RnBeads requires a specific format to match samples to IDAT files. This is first the Sentrix_ID (typically a number) and the Sentrix_Position (e.g. R03C07). If both of those columns are available, RnBeads automatically creates a column named barcode as the concatenation of those two strings. This is then also required to be the name of the corresponding IDAT files (with _Red and _Grn added). You can also directly add a column named barcode to your sample annotation.

Hope I could help you in solving the problems.
Best.

Michael
Reply all
Reply to author
Forward
0 new messages