Profile Evaluation: MS/PhD in Bioinformatics, MS in Statistics

[rtpstg673]

Hi all, I'm currently applying to a mix of PhD and MS programs in bioinformatics as well as a few MS in statistics and would love an evaluation of my profile.

Here's my background: Majored in neuroscience and completed the pre-med curriculum because thought I wanted to be a physician. After college, I decided to try something other than medicine before committing to that path, and I ended up at a startup whose core product is a matching algorithm. Quickly after joining, I realized I was more interested in data-focused work and population level impact than I was clinical medicine, so I doubled down on teaching myself python and took a few additional math and ML courses online since I had only completed calc I-III and intro stats in college. Pretty quickly I joined the (then) still-small data science team, and have continued to teach myself the relevant skills and math in the 3.5 years since I joined. For the past year, I've served as a data team lead (on the now ~15 person team) where I've juggled individual contributor work with people management, system design, long term planning etc. 

Ultimately, my intellectual interests and career goals lie in applying data science and ML to problems in healthcare. That said, I would like to buckle down on the math fundamental to the work I do everyday, so I can 1) be confident I'm using the correct tests and methods in the appropriate contexts and 2) so I can feel more comfortable reading and then implementing some of the more math-intensive papers in the field. Moreover, I've truly enjoyed the leadership aspect of my work (both the technical leadership and people management), and feel as though there's likely a ceiling in the leadership path in data science without some sort of advanced degree. So, by pursuing an advanced degree, I hope to flesh out my fundamental understanding of the math I use everyday and in the process gain a degree that will help prevent being boxed out from leadership positions in the future. Finally, for the PhD, I'm extremely interested in interpretable ML since the lack of interpretability is poised to be a major barrier to adoption of these methods in healthcare. Right now I'm leaning towards a masters (opportunity cost of the PhD is large), but if I were able to work with a group especially well aligned with my research interests in interpretable ML, I'd seriously consider a PhD.

 

----------

- School: Ivy, Major: Neuroscience (full pre-med curriculum) 

- GRE: About to take it, have been scoring 165-167Q on official practice tests.

-GPA: 3.6

- Demographic: White male from the Northeast 

-Relevant coursework: only calc I-III, and non-calc-based biostatistics course. This lack of formal math courses is what I'm worried about.

-Skills: python, SQL, a number of ML algorithms (SVM, logistic regression, various clustering, tree search, surrogate models), lots of visualization tools and packages, writing production code, writing tests, ETL (idk how relevant some of this stuff is to grad school), people management, project management

- Work experience: 3.5 years in data science at a data-product centered startup. Was an early employee at the A round (<20 people), now ~150 people post series B. I've been a team lead on the data science team for 1 of those years with 3 (and growing) direct reports.

-Academic research experience: 3 years undergrad. 4 posters at conferences, 1 published paper (3rd author, solid but not top tier journal). 

-Programs of interest: Stanford (Statistics and Bioinformatics), UW (Bioinformatics and Biostatistics), Harvard (Bioinformatics, Biostatistics, CS), UC Berkeley (Statistics). I'm looking to add some less competitive programs, so please send some suggestions! 

[DanielWarlock]

You have had no formal training in math/stats. This may not be as bad as you think for some master programs but are you sure you can handle a phd tho, given that you are admitted? You can grab a few graduate level textbooks such as theory of point estimation by Lehman, probability and measure by Billinsley.  See if you are comfortable with those. It could be a problem for someone who had no exposure to math. Remember: you will need to take courses on those in ANY PhD program and pass a test; this could be your undoing despite you being a star researcher. 

But I could be wrong. There was actually a high-school student who managed to take graduate probability class last semester from scratch. But that guy was a genius and you may be not; plus, he was an established researcher in some pure math domain, already published a dozen papers before college. So he had a lot of exposure to math already. All I'm saying is that you don't want to enrol to a PhD and have to dropout because you struggle with the classes or qual exam. It was a waste of time and money; and the struggle is bad for your mental health. You may want to do a master first at Berkeley or Harvard before applying for PhD--I see you did list those two schools and I think you will find yourself quite at home with their more "gentle" course requirement. Be cautious about going to Chicago or Stanford; the master programs there are also very hard (so I have heard)--you may also end up dropping out.

In terms of less competitive programs, you may want to apply to UF. The feedbacks from one guy who just graduated there seems to be very positive. I actually applied to only one "safe" school and it is UF. I mean, I don't want to go much lower than UF since PhD is a big commitment. But that is just me. You can check out something like Florida State--they say there could be no funding for PhD! I don't see how it will work so choose not to apply. But at least something to put on your Radar.  

But my honest suggestion is to consider doing an MBA instead of a technical degree. Have you thought about that? You have been out of school for a couple years now and haven't had much math in the first place. Now you go back to school for stats and all of a sudden are expected to do integrals, derivatives, matrices etc. Is that what you want? because that is what you will get. You don't get to use sklearn to do regression anymore. You need to actually write down things in excruciating details such as doing matrix inverse, things that are not necessary for most practical applications (though necessary for academic research). Is that truly going to help your career and your start-up? 

Edited October 29, 2019 by DanielWarlock

[bayessays]

The above suggestion that you look into a book like Lehman before starting a PhD, yet alone one not in statistics, is crazy. 

I'm not super familiar with bioinformatics admissions, but you'll need at least linear algebra to get into a decent statistics MS.

[Bayequentist]

I think DanielWarlock was a little confused. OP never stated that he wants a PhD in Statistics. Almost all of the materials taught in a typical MS Statistics program are not measure-theoretic (the probability course might touch upon a little bit of σ-algebra).

[rtpstg673]

Thanks for the thorough response DanielWarlock. Bayequentist is correct though that I'm not looking to do a PhD in anything other than Bioinformatics. Meaning, with respect to statistics programs, I'm only interested in masters. I have thought about MBAs, but imo for my path they're superfluous and a waste of $$ (and others seem to be thinking similarly with the decline in MBA apps ). 

For linear algebra, I haven't taken it for credit, but I have completed 18.06 on MIT's opencourseware. 

So, the question is: which stats MS make the most sense given my background with less formal math training? Sure I can just use sklearn all day, but one of the goals of doing this program would be to better wrap my head around the techniques I use every day and give me a solid foundation in math for understanding others. 

 

[Bayequentist]

If you can show admission committee that you know linear algebra, I think you'd have a good chance of getting into Oregon State's MS Statistics program, with funding. It'd be a good stepping stone to a better PhD program.

[BL250604]

Absolutely agree with @Bayequentist, I think you'd have a very good shot of getting funded at Oregon State. Perhaps even South Carolina (they do fund M.S. students as GA's), UVA (not sure on their funding situation with M.S. students) and UMass Amherst are all decent spots that can set you up well into a better Ph.D. program or, if you like the program, you can stay there for the Ph.D. 

[jmillar]

On 10/28/2019 at 3:58 PM, rtpstg673 said:

Hi all, I'm currently applying to a mix of PhD and MS programs in bioinformatics as well as a few MS in statistics and would love an evaluation of my profile.

Ultimately, my intellectual interests and career goals lie in applying data science and ML to problems in healthcare.

You would be fairly competitive for the Bioinformatics program at University of Michigan. About a third of the PhD students in Bioinf end up also getting a MS in Statistics along the way.

They also have certificates in areas that might interest you such as Computational Neuroscience ( https://micde.umich.edu/comput-neuro-certificate/ ) and Precision Health ( https://precisionhealth.umich.edu/about/precision-health-graduate-certificate-program/ ). Many people in the department are working on ML problems in health care or genomics and you have the option to have advisors in other departments/programs such as Neuroscience as long as they become affiliated with the Center for Computational Medicine & Bioinformatics.