Genomics and Bioinformatics?

[canidoesscience]

---

Members don't see this ad.

I'm considering applying MD/PhD to do genomics/bioinformatics work, but having some trouble really nailing down my motivation for the dual degree / considering if it's worth it.

I find the current state of genomic research extremely fascinating. My long term goal would be tease out clinically relevant information from the increasing amounts of data we generate. I'm thinking biomarkers, epigenetics, etc. I work now in relatively "low-throughput" environment doing biomarker stuff and enjoy it well enough.

The problem comes when I imagine my actual role in everything. If I were to go for it, my principle interest would be on the algorithmic/statistical side, and I'm also broadly interested (and fairly inexperienced) in HPC/Storage problems. While I've come to appreciate the intricacies of the wet lab side of the work, I have little interest in that playing a big role in my career.

Thing is, I can't come up with a good reason why your computational geneticist and physician need to be the same person. At best it seems like a medical education would be a tool for nailing down experimental design, but that's a lot of time spent for what could probably be accomplished by a long conversation with a physician. Clinical access is a pro too, but it seems like with the current trend that will be less and less of an issue.


I know I didn't bring it up, but I also really want to be a practicing physician, have wanted to since I was 5 etc etc. I waver back and forth between wanting to be the practicing physician with a strong scientific bent, and the clinically focused scientist. While the MD/PhD certainly seems like a tool to bridge that gap, I can't help but feel like it might just more be a way to delay the decision...

Any thoughts? Particularly appreciate insight from anyone in the field.

I should probably add, my major was totally unrelated in college and with the pretty stringent math requirements for a lot of programs, I think I'd need to do a masters before applying MD/PhD or PhD. I don't necessarily mind spending the time, but given the funding environment and the general outlook, I can't say I'd be happy to spend all that time to end up doing a job I could have done without all that training.

 

[gbwillner]

This is the future of medicine. You are certainly not alone in your thinking, and going into this field with an md-view of the world would make you incredibly valuable for the development of novel tools and understanding how to correctly interpret complex genomic or other clinical data.

I was totally naive in this area until the end of my PhD and was late to the game. However, I am now on the cutting edge and helping develop the landscape of genomic analyses in the clinical sphere. You could finish your training when the field is still developing and could make significant contributions to it. I cannot recommend what you are proposing strongly enough.

Good luck to you. PM me if you have any other questions.

 

[Dr.Tao]

"This is the future of medicine."

+1

 

[haanman]

Nice post OP, it's clear you've thought about this a lot. I'm doing my PhD mostly in genomics/bioinformatics, and I also think it's a great way for an MD/PhD to go. The PhD changes your clinical thinking too; for the most part, my former classmates just haven't had the chance to engage with this side of medicine.

If you were interested *just* in the algorithms/statistics and/or were considering moving out of biomedical science altogether and into tech/data science/whatever, a straight PhD in comp sci/applied math/stats might be a better investment of time. But since you really like the clinical side and want to practice, why hesitate? MD/PhD programs were meant for people like you.

About the master's: I did one in math before starting med school, because I couldn't decide whether to try to be a mathematician or a doctor. I don't regret it, but you shouldn't need it as long as you choose your MD/PhD program well. A place with a high concentration of bioinformatics-engaged PIs and thoughtful training plans for grad students should give you everything you need.

 

[The Skeptic12]

well the level of math you speak of in bioinformatics is way below what people in wallstreet hedge funds use for their probalistic models. If you used any of the models or nomograms you see in biomedical or medical or bioinformatics research to trade with you would lose your shirt on wallstreet. The best applied math school in the USA is Courant at NYU math finance. The only people who are at their level (maybe) are those at the Princeton Inst for advanced study. the problem is I am an engineering student who has a Phd in Exp Psych and I know the real mathematicians and so far big data, massive parallel computing and complex probabilistic models used on wallstreet and in weather forecasting have not found their way into medical science. Understanding complex processes like aging and cancer will require complex models that are multidimensional and wont work at first. No one is approaching this the way it should be. Its as if you have to decode a machine more complex and less deterministic than an Enigma machine with no computers, not enough data and people who cant even call themselves mathematicians compared to those who worked with Turing at Bletchley Park. How far do you think they will get? All they will do is do 1 genome drugs that provide little more than a bandaid for something like metastatic cancer. Real time monitoring and data on many levels from a complete genome are they way to go.. We havent done it yet.. what would it take.. just money and vision, something lacking here where we use Mice and have arrogant physicians who dont know what they are doing.. all they do is say we havent made any progress ie Alzheimers, Prostate Cancer etc. Well yes not the way we are doing things now!

This is really frustrating for me. I work hard at my math courses and programming (I work on Wallstreet) and I look at a mess that MDs/Phds make of research.. we have a long way to go unless we have a president who isnt a lawyer or a billionaire who knows what is going on... I see us going nowhere. I suspect though that the military ie Airforce might be doing what needs to be done based on my reading of their other amazing research ie antimatter. I have no security clearance but imagine the power you would have if you could stop any cancer or ****** aging? Sound like science fiction or paranoia? Just look at what Snowden found. What would be worse is if civilians cede this knowledge to the military.. we had better get on it and fast. See what it takes to get in math/finance at NYU, I know someone who taught there and she tells me about all the people with PHds ie from MIT who just have real trouble with the math.. everyone does at first. This is not to criticize but to tell you the real deal. Get a real math degree at a top place.. preferably in something like finance.. these are the folks that can help the MD's and let me tell you they are like blindmen feeling the tail of an elephant.

We are in trouble here and I am writing my first article as fast as possible but hey I have to make a living too!!! Glad that you realize that this is the future but keep in mind that complex math requires years of struggle to get unless of course you are Alan Turing or..........

Good Luck.

 

[halcyonpage]

I did bioinformatics for a couple years at a start-up. I actually found it dreadfully tedious and boring, probably because we had so few resources and had to write our own scripts and use freeware. I think you're lucky since you really love bioinformatics and math. I just like biology lol. And your stats are super impressive, I wouldn't even worry unless there was *1* dream school you just HAVE to go to haha.

When I was doing it, I wish I had more information/context about the cases the data came from. I think if you get the MD as well as the PhD, you will have a better understanding of how the disease state translates into symptoms, how to better group the data you get from a diverse population, and whether certain results will be directly or indirectly related to the mechanism you are studying. Personally, I like to rely as much on consultants as much as my own common sense. Talking to patients might also illuminate otherwise hidden genetic traits that could confound your data.

 

[MindCastle15]

This is the future of medicine. You are certainly not alone in your thinking, and going into this field with an md-view of the world would make you incredibly valuable for the development of novel tools and understanding how to correctly interpret complex genomic or other clinical data.

I was totally naive in this area until the end of my PhD and was late to the game. However, I am now on the cutting edge and helping develop the landscape of genomic analyses in the clinical sphere. You could finish your training when the field is still developing and could make significant contributions to it. I cannot recommend what you are proposing strongly enough.

Good luck to you. PM me if you have any other questions.

Based on ur reply, may I ask if you got an M.D. degree later on? Does it actually help, in terms of research?

 

[gbwillner]

I did get an MD concurrently with the PhD. It helps for research in disease-related projects, of course.

 

[MagicBlueSmoke]

@gbwillner, do you agree with what @the skeptic 12 wrote about biostatistics/informatics falling way behind the techniques used in finance? If so, what's a good way to learn those techniques in an MDPhD setting? As an undergraduate, I took many systems engineering courses geared towards undergraduates going into financial engineering or graduate students in finance programs. As far as I can tell, it is the same sequence as people going into the bioinformatics track for biomedical engineering, biostatistics, etc. What are your thoughts?

 

[gutonc]

@gbwillner, do you agree with what @the skeptic 12 wrote about biostatistics/informatics falling way behind the techniques used in finance? If so, what's a good way to learn those techniques in an MDPhD setting? As an undergraduate, I took many systems engineering courses geared towards undergraduates going into financial engineering or graduate students in finance programs. As far as I can tell, it is the same sequence as people going into the bioinformatics track for biomedical engineering, biostatistics, etc. What are your thoughts?

We all completely agree with random new poster who dropped in here to s*** on all of us and then leave. You should too.

 

[gbwillner]

Since I am not in financing, I can't speak for the complexity, importance, or interest in their mathematical modeling. I'm satisfied with trying to understand the world we live in and attempt to cure disease, and do not have lofty goals like trying to figure out a way to cheat people out of their retirement funds.