Microarray image explanation

Question

[n.lu]

While analyzing my microarray data, I looked at the images of the chips. It has an empty band on the right side and a spot in the middle - photos of the first array are attached bellow. The same pattern is present on all (16) microarrays in my experiment. Raw image: PLM image: I dont have any experiance looking at chip images, so the band looks confusing to me. Is that normal or concerning? Anyone has an explanation?

Thanks in advance!

[h.mon]

What is the platform?

[n.lu]

GeneChip™ Mouse Gene 2.0 ST Array, Affymetrix

Answer 1

[Kevin Blighe]

Hey, I don't see any attached figures to your question. However, I have analysed many different types of microarrays and I always check the chip images just to look for any 'major' artifacts/errors.

If the bands that you describe are very consistent across the chip images that you've got and have very defined shapes, then they are most likely the regions where the negative and positive control probes reside. The central part could be the part of the chip where the sample is actually loaded.

I've attached my own example where you can see very defined regions that are clearly marked - these are consistent across all of my chips from this particular study. I decided to highlight blue-red just to be different.

[n.lu]

Hi, Thanks for your answer! That explains the central square, thanks! I will try to reattach the images so you can see the band on the right side (bot images of the same array, one is PLM).

[Kevin Blighe]

Hi m.lu. Thank you for posting the images of the chips. Luckily, I have data from that same design version from Affymetrix, but for human. I see the same large band at the right, and I can only conclude that it's the control probe area. Something at the back of my mind is telling me right now that Affymetrix decided to include a lot more control probes in latest designs. In older Affymetrix chips, you won't see such banding because the control probes were used different - in those, you'll just see black/white dots like a broken television.

So, I would say that your microarray chips are just fine. Also look at the boxplots of raw and then normalized counts. There is one part in your first array that looks unusual but it may just be a genuinely highly expressed gene. For example, if it's blood mRNA, then the highly expressed transcripts could be from the haemoglobin sub-units genes.

[n.lu]

Great, thank you very much for your answer! The darker spot on the first array is not present on the others though.. I will attach the boxplots from before and after RMA bellow for you to see. They look fine to me, but id like to make sure.

[Kevin Blighe]

Yes, looks okay to me! I would be happy to analyse those!

[n.lu]

Thanks for all the help!