DESeq2 Design Formula
Entering edit mode
gkuffel ▴ 10
Last seen 3.1 years ago
Loyola University

Hello community,

I have posted this question before but wanted to see if I can gain additional clarity. I have an miRNA experiment with the following parameters:

SampleID     Treatment     Timepoint     Patient

   P1_0                  TB                 0             1

   P1_1                  TB                 1             1

   P1_2                  TB                 3             1

   HC1_0                None            0              HealthyControl1

   HC1_2                None            3              HealthyControl1

   P2_0                  TB                 0              2

   P2_1                  TB                 1              2

  P2_2                   TB                 2              2

   P2_3                  TB                 3              2

   HC2_0                None            0              HealthyControl2

  HC2_3                 None            3              HealthyControl2

My questions are:

1. I don't have replicates for the time points is this a problem (the investigator was on a budget)?

2. How can I test for differences between patients?

3. How can I test for differences within a patient (as a result of the treatment)?

4. How can I utilize my healthy controls? Each patient has age-matched healthy controls. 

Much gratitude for any help offered.

DESeq2 michael love • 361 views
Entering edit mode
Entering edit mode
Last seen 13 hours ago
United States

If you posted before it would be useful to link to the other post (and back) so that other users can navigate the threads better.

This is a difficult design for fixed effects model, because you are missing a few time points for various patients and only have two patients per treatment. I'd say the easiest approach would be to model the patient effect in limma with duplicateCorrelation, and then focus on the treatment and time point interaction. Take a look at the limma voom guide with the use of duplicateCorrelation.

Entering edit mode

Thanks Michael. I will add the link. Also I actually have 36 patients that received the treatments and two healthy controls per patient, I was just trying to show a simplified version. So your advice is to use Limma instead of DESeq2?

Entering edit mode

Well you got the advice based on what you posted, which didn't mention additional data. Given that you have missing time points for some of the patients, duplicateCorrelation will be a simpler approach in my opinion. It's harder to lay this out with fixed effects to estimate the treatment specific time differences when some samples have missing time points.


Login before adding your answer.

Traffic: 212 users visited in the last hour
Help About
Access RSS

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6