2 replies to this topic

[vetticus3]

Hi,
I was wondering if anyone has any views on microarrays used in diagnostic purposes.  I have to do product development proposal as part of my degree, and since the area i'm researching is pretty limited in developing a product, i (read 'my supervisor') got to choose something else, and I (my supervisor) got this great idea of using a microarray to test breast tumour samples for a two gene expression ratio.  
I think microarrays make really pretty colours, and there's a really big book with lots of pretty pictures on my desk... yeah, anyway.     The idea of developing a microarray to test the expression of two genes that may influence the way cancer is treated seems pretty interesting.  Unfortunately, i don't think it would really take off.  I've got this idea that microarrays are really expensive and big, and difficult to understand...pretty colours.  
What do you think about microarrays for diagnostic purposes?  If you could change them in anyway, what would you do?  
Don't be constrained by logic and the actual limitations that these machines have.  This is a perfect world...    Complain about their limitations!!!  ie.. Make the reader in pastel pink, and the box of reagents comes with a free bottle of perfume....and a magic monkey that makes coffee... or that uses FFPE samples, and has a CD that explains those pretty colours.
cheers,
vetticus

Edited by vetticus3, 28 September 2004 - 09:00 PM.

[uaue]

yep it´s quite expensive. it has it´s advantages though like you can impress everybody with the sophisticated analysis and of course the colors (well the really important ones are green for down-regulated and red for up-regulated).but mind you it can be frustrating as well specially when you have 90,000 genes to look at! and the clustering...pufff i don´t want to talk about it.but if you have the funds go for it!and good luck

[vetticus3]

This is a completely theoretical product... it's just for an assignment, but it has to have sound science behind it   .
If you were a... those people who look at tumour samples all day and decide what kind of tumour it is, stage, etc... pathologist (?), and a very perky sales man/woman  with lots of freebies comes into your office and says "i've got this wiz bang microarray that will read your FFPE samples for the expression of two specific genes, that'll help you decide whether that cancer patient will resond to drug X, or drug Y. Very snazzy..."  what would your expectations be?
Would you think, get out of my office, it'll cost too much, i'm colour blind so those colours suck...
Would you want the array to come with it's own reader, or able to be read with any generic reader?  
How about a guarantee that clustering doesn't take place?  And a computer program that analysis the data for you, and just spits out the answer "use drug X".