# Issue with DGE analysis using ds.limma

**URL:** <https://datashield.discourse.group/t/issue-with-dge-analysis-using-ds-limma/1119>\
**Category:** Developer support\
**Created:** [15 October 2024 10:31 UTC](https://datashield.discourse.group/t/issue-with-dge-analysis-using-ds-limma/1119 "2024-10-15T10:31:16Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![FrancescoAnzillotti](https://avatars.discourse-cdn.com/v4/letter/f/6de8d8/32.png) [@FrancescoAnzillotti](https://datashield.discourse.group/u/FrancescoAnzillotti)\
**Post date:** [15 October 2024 10:31 UTC](https://datashield.discourse.group/t/issue-with-dge-analysis-using-ds-limma/1119/1 "2024-10-15T10:31:16Z")

</div>

Hi, I’m having some issues using the ds.limma function. Specifically, when following :

> **[14 Differential gene expression (DGE) analysis | Orchestrating...](https://isglobal-brge.github.io/resource_bookdown/differential-gene-expression-dge-analysis.html)**
>
> 14 Differential gene expression (DGE) analysis | Orchestrating privacy-protected big data analyses of data from different resources with R and DataSHIELD

my output is :

```r
"Command 'limmaDS2(rse, res, 2L, NULL, 2, NULL, FALSE, \"none\")' failed on 'study1': 
Error while evaluating 'dsOmics::limmaDS2(rse, res, 2L, NULL, 2, NULL, FALSE, \"none\")' -> 
Error in dsOmics::limmaDS2(rse, res, 2L, NULL, 2, NULL, FALSE, \"none\") : \n object 'Set.counts' not found\n"

```

Troubleshooting:

- I locally downloaded the dataset used in the guide above
- I downloaded the code of the ds.limma function (client side) and limmaDS function (server side) available on [dsOmics/R at master · isglobal-brge/dsOmics · GitHub](https://github.com/isglobal-brge/dsOmics/tree/master/R) and combined them together to run locally on the downloaded dataset.

Running locally produces the expected output (the analysis is correctly executed).  
What seems to be the problem is the connection between the two functions when operating in client-server mode, specifically this piece of code when calling limmaDS

```r
Set <- eval(parse(text=Set), envir = parent.frame())

```

---

<div class="post-metadata">

**Author:** ![xescriba](https://avatars.discourse-cdn.com/v4/letter/x/ee59a6/32.png) [@xescriba](https://datashield.discourse.group/u/xescriba)\
**Post date:** [24 October 2024 15:00 UTC](https://datashield.discourse.group/t/issue-with-dge-analysis-using-ds-limma/1119/2 "2024-10-24T15:00:13Z")

</div>

Dear Francesco,

Thank you for raising this issue. I have taken a look at the code and traced back the error.

Please note it has been fixed, see commit [fix: type not used here · isglobal-brge/dsOmics@9f5fe8c · GitHub](https://github.com/isglobal-brge/dsOmics/commit/9f5fe8c6c0acfdb8dcb8da6f451458107e64b0fa).

I have just tested to install this latest version on the “omics” profile of [opal-demo.obiba.org](http://opal-demo.obiba.org) and sucessfully run.

Also, seems the document you link should be updated so the builder looks like this

```auto
builder$append(server = "study1", url = "https://opal-demo.obiba.org", 
               user = "dsuser", password = "P@ssw0rd", profile = "omics",
               resource = "RSRC.tcga_liver", driver = "OpalDriver")

```

Anyway, let me know if it is also solved on your end. On the meantime I will get in contact with the maintainer of opal-demo so the latest version of dsOmics is there permanently.

Have a great day,

Xavier.