Usefulness/career path of MD/PhD with PhD in genomics/bioinformatics?

[shinny]

Part of me thinks that a bioinformatician is more research support, like a faculty biostatistician that collaborates with multiple PIs at his/her institution. It’s a field I’m interested in but haven’t been able to find much in terms of a use case for an MD/PhD and the kind of career one would have. For context I’m entering an MD/PhD program.

1) Is PhD training in bioinformatics useful for a physician? Is there a growing need for these kinds of physician-scientists?

2) What kind of career would one have? Are there physician scientists trained in bioinformatics heading their own labs and receiving NIH grants? Would it be possible to have a lab that both heavily uses genomics tools and does bench research based on the results of the genomics experiments, while being a physician at the time? Or would this require a postdoc to gain the necessary lab streetcred (pubs)

3) What specialties might work well with this training? It seems almost omnirelevant barring perhaps surgical specialties.

---

Members don't see this ad.

 

[tortuga87]

1) It will not be useful for day to day activities of a physician. As far as research, this depends on the type of research questions you want to answer. There is no large need for physician scientists who are well versed in informatics. In other words, people will not beg for your services and pay you lots of dough, but they will ask for your free labor to analyze their datasets and give you a pat on the back because "they were never good with numbers."

2) The career path depends on how you set it up. You can be that support guy/gal, or you can be the person who does his/her own thing. It really is up to you. There are many different paths after residency (or after MD/PhD, if you decide to forego residency). Like all other sciences, the best bioinformaticians are the most versatile in the sense that they can develop fancy methods and they can verify the results with experiments. Being able to run experiments is very important for writing solid arguments in papers because it is much easier to elucidate causal effects from experimental data than observational data. If you cannot run experiments, you essentially just end up writing papers which go along the lines of: "I have this fancy method. It verifies past results. Go to this URL to use it (please)" as opposed to actually advancing biological knowledge with the method.

3) Most MD/PhDs in this area practice medicine and do minimal research because clinical practice pays much more. On the other hand, most MD/PhDs who do majority research do not practice medicine because those who shoot for a serious research career realize that their clinical component demands too much time away from their research. The most popular subspecialty is probably heme/onc for obvious reasons. People who do imaging informatics by far go into radiology, or pathology second in line. The other non-surgical specialties that traditionally attract MD/PhDs get a trickle. Virtually none in the surgical specialties (this is more of a bioengineering/device design route). You will probably be one of the few MD/PhDs in informatics at your university regardless. At the end of the day though, it does not matter what other people go into because you will probably have a vision of what you want to do.

 

[kchan99]

Hard to tell where the genomics/bioinformatics market will go, but depending on specialty/subspecialty selection, IMO bioinformatics can be helpful. For example, clinically, in molecular pathology or medical genetics, you could run a clinical genetic testing section. with overseeing a whole exome/genome group. If you end up having a physician-scientist career and don't have your own grant, you could get protected time funded through research collaborations providing % effort with salary support. It depends who else learns computer programming in the future.

I'm in image analysis, but I'm not a radiologist, pathologist or biomedical engineer, but I dabble in both radiology and pathology. I have minimal experience in bioinformatics, but if I had time, moving to bioinformatics, isn't too difficult. I would just have to learn new languages.

 

[shinny]

Thank you both very much for the input. I have two other questions-

1) If pursuing a PhD in bioinformatics instead of the biologies, how would one be able to verify these results experimentally? Where/when would the typical lab training be acquired to run a lab?

2) I've read conflicting accounts- is a bioinformatics/CS PhD typically longer or shorter than a biology PhD? I am sure it varies, but wondering if it swings one way or another and in what situations

 

[tortuga87]

Thank you both very much for the input. I have two other questions-

1) If pursuing a PhD in bioinformatics instead of the biologies, how would one be able to verify these results experimentally? Where/when would the typical lab training be acquired to run a lab?

2) I've read conflicting accounts- is a bioinformatics/CS PhD typically longer or shorter than a biology PhD? I am sure it varies, but wondering if it swings one way or another and in what situations

1) Ideally you would work in a lab that does both "computational" and "wet lab" work. If you cannot work in such a lab, then you will have to find ways to get the experimental training by constructing your own agenda, since your PI will most likely not guide you through this process. You will not be expected to design experiments as well as a molecular biologist, but you should be able to perform a rough start (e.g., be able to design and run experiments using flow cytometry, western blots or whatever common tools are used for the disease/biological area you are interested in). Note that experimental training is not required, but I think we can all agree that experiments are important to create better arguments for your discoveries.

2) Bioinformatics PhDs take slightly less time on average. You have probably heard of the argument that computational PhDs dont require you to wait for things to grow in the lab, so the PhD takes less time. Informatics also rarely involves busy work because you can usually just write some code to solve the problem (just don't get caught up in a data cleaning project btw). However, the learning curve is generally steeper for CS/statistics type work than biology because more background knowledge is required. This means that you will also be taking many classes which will seem unusual to your colleagues doing wet lab work. I was personally not a fan of trying to finish the PhD as soon as possible. I instead approached the PhD as a very special protected period in my life where I could dedicate nearly 100% of my time to becoming the best researcher possible (this will probably never happen again during my career). That meant taking many extra classes and working on many extra projects. MD/PhDs who focus on graduating early by focusing almost exclusively on their thesis can realistically finish in 3 years (some even do it in 2). Otherwise, 4 years is a reasonable length of time for MD/PhDs.

 

[ellifino]

So my PhD is in genetics but I have an IT background, a math degree, and my genetics research was all bioinformatics / big data stuff, so i consider myself a bioinformaticist. I'm now in an anesthesiology residency. I felt like my background was very well recieved during the application process, and I know have more research offers than i know what to do with. Its a rare skill to have, and highly in demand, at least in my experience. Feel free to PM me if you have more questions!

 

[maxknux]

I was thinking about this route but I am trying to see if this would be viable career path. It looks like people go in either Clinical or Research Heavy.