Entering edit mode
Dipro
•
0
@dipro-13570
Last seen 7.4 years ago
Hi,
I was wondering if I could get some advice on controlling for random effects using DESeq2. For example, given simulated data:
| Treatment | Batch | Estrus |
| Control | 1 | 1 |
| Control | 2 | 2 |
| Control | 3 | 1 |
| Control | 4 | 1 |
| Control | 1 | 2 |
| Control | 2 | 2 |
| Control | 3 | 2 |
| Control | 4 | 1 |
| Control | 1 | 3 |
| Control | 2 | 1 |
| Treated | 3 | 3 |
| Treated | 4 | 3 |
| Treated | 1 | 2 |
| Treated | 2 | 3 |
| Treated | 3 | 2 |
| Treated | 4 | 3 |
| Treated | 1 | 2 |
| Treated | 2 | 1 |
| Treated | 3 | 1 |
| Treated | 4 | 2 |
What is the best way to control for batch and estrus? I understand that DESeq2 does not support random effect analysis. As such, I was wondering if I could be advised on a work around for this type of analysis?
Thanks
One potential workaround might be to adjust for these factors using a combination of design matrices perhaps including batch as a covariate in your model. Alternatively, considering other tools that allow for random effects, such as limma, could also provide more flexibility.
By the way, have you checked out Crazy Cattle 3d? It is a fun project that showcases the merging of creativity and technology in farming. It might not directly relate to your RNA seq question, but it's fascinating to see how different fields are innovating. Looking forward to seeing more of your work!