Hacker News Reader

Comment by dekhn | Hacker News Reader

So, as cancer is one of the leadng causes of death, your prediction implies that population (especially older population) will increase. Or people will die of other things. That's an important thing to be prepared for (the other thing people die of in first world countries is heart disease).

Anyway, I suspect the future won't be quite so rosy as you predict. Many people have predicted revolutions in cancer treatment, but realistically, we've only made incremental (great, but only incremental) progress.

View on HN

Have you read The Emperor of all Maladies? The impression I got from the book was that we are finally beginning to understand the mechanisms behind specific cancers and create therapies to fix them. It seems like we are close to beating cancers one at a time with targeted drugs.

Yes I have read it. PS: I'm a scientist and studyng cancer is part of what I do. Please read "Molecular Biology of the Cell".

yes, we're learning a lot, but no, the promised revolutions aren't forthcoming. Not only are there technical reasons why it's unlikely we'll see a very near term revolution in treatment. There are huge political, monetary and bureaucratic reasons, too. They are just as important to understand as the technical reasons.

From the technical side of things, where do you see the biggest challenges today? Where do you see the biggest potential wins? What are we not looking at deeply enough?

This is a field that deeply interests me, but my knowledge is effectively zero. I'd love to get your insight, as one that is on the ground in the field.

from my recent Human Genetics course, I learned quite a lot about what cancer actually is. It helped a great deal to understand why there's no "cure" and treatments are usually brute force and imprecise.

Cancer is basically a generic term for the loss of a cell's division control cycle - the cell starts to divide uncontrollably, forms tumors, metastasizes, causes neovascularization to feed itself, etc. This loss of control can happen for so many different reasons related to the quantities and ratios of proteins synthesized that each cancer is like a separate class of disease that needs its own treatment.

It's the same reason that there's no single fix for "car is broken". Except our understanding of this car is orders of magnitude more limited than one that we designed ourselves.

I'm very much interested in this as well.

"There are huge political, monetary and bureaucratic reasons, too."

As a standard web / mobile developer how could I help at least in those areas?

On the technical side, I think we need to master data integration across many sources. For example, health records, sequencing, and clinical trials all contain tremendous amount of data, but combining them and generating knowledge is very hard. Much of this is just translating what web companies have developed, and understanding how to make it palatable to health researchers, patients and companies. That's partly technological, but honestly, the technological problems are pretty small; we could make a tremendous forward progerss in cancer if everybody's health records were fully digitalized, normalized and harmonized, and available to cancer researchers (even ones not doing trials) for aggregation and analysis. This is one reason I've donated my genome to the public domain- but we need millions more to do this.

Also, the way we fund and execute most health research is very broken. We could organize and manage this far better than we can- but it would take a lot of changes and stepping on a lot of toes. Again, there's a tech side, but the tech side is "easy" compared to the politics and money parts.

Reply on news.ycombinator.com