[Question]: Should one Re-run DESeq after removing genes with low values

[tamuanand]

Hi @federicomarini

I happened to find this behavior on the "Counts Overview" tab. Let's say I use "airway demo" data and "filter DDS by threshold on row sums of the counts" and choose 10 the first time

• the resulting table shrinks to 22,008 entries
• If I now change the threshold to 8 (or anything <10), the number of rows in the table does not increase (it remains the same at 22,008). Is this expected?

Does one have to re-run Step 3 in the Data Setup tab (run DESeq) after doing the above filtering (say using 10 as threshold with row sums)? I ask so because the ideal userguide has this - https://federicomarini.github.io/ideal/articles/ideal-usersguide.html
Additionally, [ideal](https://bioconductor.org/packages/3.20/ideal) has an option to include a filter step at the gene level by removing genes with low absolute or averages low values. After this, it might be possible to have to re-run the analysis in step 3 from the Data Setup panel

Thanks in advance.

[federicomarini]

After changing the threshold, you should still "re-create" the dds object in use for the app.

In theory, you should run DESeq() on the features you keep in your object.
I did not run any benchmark on what you obtain for the retained features if you keep the original values from the "full set" vs the re-done one, but I am expecting them to not vary too massively.