Question: Can remote, but same type, tissue be used as RNA-Seq normal control for tumor study?
gravatar for CY
3 months ago by
United States
CY270 wrote:

Usually, adjacent tissue is used as RNA-Seq normal control. However, Adjacent tissue is not accessible in some case. Can I use remove, but same type, tissue instead?

Beside, adjencent tissue in many cases is not typical 'normal' tissue. It may contain certain degree of tumor cells and immune cells which may bring bias to differential expression analysis

rna-seq • 156 views
ADD COMMENTlink modified 3 months ago by kristoffer.vittingseerup1.4k • written 3 months ago by CY270
gravatar for kristoffer.vittingseerup
3 months ago by
European Union
kristoffer.vittingseerup1.4k wrote:

In principal yes - you would just have to argue why you would use something else (and you might need to prove the second hypothesis - for most studies the doctors are very carefully with that kind of thing).

In practice this is a lot harder because of batch effects. When a RNAseq experiment is done there is a large effect of the specific day, person age of chemical, protocol used etc - this effect is revered to as a batch effect. This mean that if the samples you want to compare was done in multiple batches you need to correct for this using batch correction (or building more elaborate models).

In your case it means that you cannot take a set of samples from one condition in one study and compare it to samples from another condition in another study because you can never distinguish what is the true effect and what is the batch effect. So in practice no.

If not properly handled this is what happens.

ADD COMMENTlink written 3 months ago by kristoffer.vittingseerup1.4k

Thanks for the comments. I am aware of batch effect in RNA-Seq analysis. We will extract samples in the same batch.

My concern is the use of rmote normal tissue as control instead of adjencent tissue which is not accessible for some reason. I imagine that would be all right since the same tissue would have similar expression profile even it is remote from the tumor. What do you think of that?

ADD REPLYlink written 3 months ago by CY270

I would generally trust samples obtained from adjacent tissue because:

  1. From data originating from micro-dissections and single cells from ajecent tissue the amount of tumor cells is very small (<1%) whereby for bulk data it does not matter
  2. I have never herd of anyone suggesting adjacent samples in fx TCGA are contaminated (and those samples have been analyzed by a lot of people
  3. I believe that the surgeons are doing a good job staying clear of the tumor
ADD REPLYlink modified 3 months ago • written 3 months ago by kristoffer.vittingseerup1.4k
Please log in to add an answer.


Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 2.3.0
Traffic: 1462 users visited in the last hour