Gene Chips to the Rescue

Bamafan77 writes "USAToday has a pretty interesting read on Joseph DiRisi and how his viral gene chip (a slide that hold large arrays of genetic material)helped id the cause of the SARS outbreak. Gene Chips (or more properly, microarrays) are an interesting a multidisciplinary application of programming, biology, and hardware."

*crunch*

[Michael.Forman]

Viral gene chips, betcha' can't eat just one.

Michael. [michael-forman.com]

Re:What's going on?

[gene_tailor]

No, gene chips are amazing actually... Maybe there's a special pr0n movie on right now?

Gene Chips

[gene_tailor]

This is a nice and useful, but predictable, way to use "gene chips" (DNA microarrays). As usual there are some inaccuracies in how the press talks about science though... Here are a few comments on the article from my point of view, as a professional molecular biologist:

1) "Gene chip helps identify cause of mystery illness" --this is only sorta true. The chip data suggests that the samples from the SARS victims contain something that is related to coronaviruses. However, the gene chip alone cannot tell you what something is if it is new, only that it's close to some of the 1000 (in this case) known things on the chip. As in, this is a four-legged mammal like a horse or pig; doesn't tell us it is a cow, just that it's similar to other four-legged animals in some ways. Also, finding the same thing in multiple samples is suggestive, but certainly not conclusive proof that this is the cause of the illness! It could be a big coincidence that they all have it, and the article doesn't even tell us how many samples were tested.

2)Article states "[DeRisi] knew he was dealing with something the world's scientists had never seen before. " --No, the results just show that the samples are not one of the 1000 on the chip, not that this virus has never been seen elsewhere in any hospital or lab....

2) " a corresponding boom in microarrays containing those nonhuman genes hasn't occurred." --On the contrary, microarray tech is a hot topic in every organism that has a sequenced genome, and in some where they only have partial sequence as well.

Re:Gene Chips

[Anonymous Coward]

"'Gene chip helps identify cause of mystery illness' --this is only sorta true. The chip data suggests that the samples from the SARS victims contain something that is related to coronaviruses. However, the gene chip alone cannot tell you what something is if it is new, only that it's close to some of the 1000 (in this case) known things on the chip ... Also, finding the same thing in multiple samples is suggestive, but certainly not conclusive proof that this is the cause of the illness!"

This criticism

Re:Gene Chips --PNAS link

[gene_tailor]

FYI: the DeRisi lab has a copy of their Nov2002 PNAS paper online here [ucsf.edu]. Interestingly, according to this paper their chip has 1600 oligos from "~140 distinct viral genomes".

Over/under glorified?

[spotted_dolphin]

I'm not particularly fond of the tone in which this article was written. I think that it's terrific that such a recent (8 years) technology has taken part in such a big news event as SARS; however, there is a significant lack of information provided about the technology itself.

I am happy to be one of many users of this DNA chip/microarray technology and this article has terribly simplified its uses and the findings thus far. DeRisi is looking at 1000 sequences, but you can easily place 20 000 sequences on a chip. That's a lot of information! Not to mention, a lot of computing power is being put into storing, let alone analysing all the data generated.

So, we're down to needing the help of bioinformaticians to sort through and organize the 'laser microscope'--essentially scanner--output, but as biologists we still have to figure out what the story being told by this number is. If we're looking at the genes of an entire organism we have to sort out what pathways are involved to generate a certain biological state (ie. cancerous vs. normal) and whether it's significant or not. There's so much we don't know and entire thesis projects are devoted to analysing such data.

DeRisi's identification that the SARS virus was 'different' is merely the tip of the iceberg of the microarray technology.

Re:Over/under glorified?

[JGski]

> I'm not particularly fond of the tone in which
> this article was written. I think that it's
> terrific that such a recent (8 years) technology
> has taken part in such a big news event as
> SARS; however, there is a significant lack of
> information provided about the technology
> itself.

From the picture and the array density quoted it looks like they are spotting their own slides.

The article is wrong about their being no virus microarrays: Affymetrix has had an AIDS array for

Affy chips

[spotted_dolphin]

Affymetrix chips are photolithographic oligos and *very* expensive. We've completed some studies with their chips, but we mostly stick with our own custom printed cDNA arrays. Our human microarray has around 10 000 sequences and density isn't a problem. We don't have to pay the license fee because we aren't selling them. Plus, we don't have to use the incredibly expensive Affymetrix machines nor their incredibly expensive data software.

More importantly, only one mixture of RNA/DNA can be analyzed on a

Re:Great story, if it was true

[spotted_dolphin]

Actually, we now have the technology and computing power to get a pretty good map of a virus in well under a week.

Understanding how the virus works (infection mechanism, etc.) and what it does to its host is completely another ballpark.

yawn

[BortQ]

Call me when they come in sour cream and onion.

A Bit more Detail

[NeuroMorphus]

I agree that this article didn't provide many details about the project. But according I've read other articles about Joseph DiRisi's research, and apparently the software used for the project was writting in Python

Big plus to Python's Numeric module.