First comprehensive analysis of two cancer genomes | Hacker News Reader

First comprehensive analysis of two cancer genomes

eurekalert.org

34 pointsepi0Bauqu11 comments

Pooh BBC, linkbaity headline and no link to the study? Well, they're really bad about external links anyway...guess they don't want people leaving their website for any reason grumble

Interesting news though, thanks. More informative press release: http://www.eurekalert.org/pub_releases/2009-12/wtsi-lca12150...

Project page with lots of resources: http://www.sanger.ac.uk/research/projects/cancergenome.html

Nature journal articles in question (you'll almost certainly need a university account to see these):

Melanoma: http://www.nature.com/nature/journal/vaop/ncurrent/full/natu...

Small-cell lung cancer: http://www.nature.com/nature/journal/vaop/ncurrent/full/natu...

Erin Pleasance is the first author on two Nature articles in the same issue. That is insane!

They didn't really crack it. They found a whole bunch of dna errors in cancer, but they don't know which ones matter.

They don't even know which are DNA mutations, and which are systematic sequencing errors.

Correction: two cancers, not all cancer.

BTW, don't downvote the preceding comments complaining about this post. They're responding to the article that was here before I replaced it with the one anigbrowl suggested.

Would it have been better to simply submit a new article?

Would the idea now be to compare multiple cases of lung cancer to identify the genetic error that causes the tumour formation and develop a way to actively target said errors?

I mean would it eventually become feasible to put a chemical (or something) into a cigarette to greatly reduce the risk of lung cancer? Similarly could a chemical be devised for addition to sun block, or even after-sun to actively fight the cancer before it develops.

Chemical as in simple compound? No. I have no idea what the worst mutagen in cigarette smoke is but there are lots and their effects are going to be random statistically, sometimes they'll change this gene, sometimes the other, though some sections will be especially vulnerable/likely to mutate in a specific way.

I could see some bloody complicated RNAinterferase thing working as a targeted anti-mutagen, in theory. But if you can deliver an airborne, persistent heat tolerant RNAi cancer is at worst a chronic disease like diabetes is now.

Could you diff a healthy genome and several cancerous genomes and come up with the offending commit? Can someone give me more background on the ins and outs?

Well, in the article it says there are 30,000+ mutations, so narrowing it down to the specific ones caused the cascade of other mutations is far from trivial.

A VC system with a corrupted database silently wrecking everything that touches it might not be a bad metaphor, though. (I've had it happen with Perforce. Not fun.)

Reply on news.ycombinator.com