PhD Admissions Retrospective Post (Computer Science, Computational Biology, and Genome Sciences Programs)


In Fall 2022, I applied to PhD programs in computer science, computational biology/bioinformatics, and genetics/genomics. To celebrate the end of my admissions cycle, I put together some general advice and thoughts for future applicants. This is mostly aimed toward people about to go through the application process—if you still have a few years left, my only advice is to get as deeply involved in research as possible.

This is based on my own experiences, my advisors’ advice, and all of the admissions-related content I absorbed over the past few years. The usual caveats apply: there are exceptions to every statement, I am not an expert on admissions, this is just my own perspective, and some parts may not be applicable to all subfields. Also, I primarily applied to programs in the US (plus a couple of programs in the UK), and the process can be very different in other countries.


Timeline

If you have no idea what to expect in terms of timeline, here is a rough sketch of what the admissions cycle might look like.

September to October: Many applications and application assistance programs open.

October: Many fellowship applications due.

around December 1: Many biosciences applications due (be sure to check the actual deadlines).

December 15: Many CS applications due (be sure to check the actual deadlines).

mid-December to early February: Programs send out interview invitations.

January to February: Interview weekends (for biosciences programs) and brief online interviews (for CS programs).

February to March: In-person admitted students weekends (for CS programs, as well as biosciences programs that had online interview weekends).

April 15: Deadline for you to respond to funded offers.


What do admissions committees care about?

I won’t spend too much time on this topic, since there are many guides written by more qualified people about what admissions committees care about. But it’s worth emphasizing: the most important factor in admissions is evidence of ability to carry out research, as reflected through past research experiences/outcomes, letters of recommendation, your statement of purpose, and perhaps notable and relevant extracurricular activities or awards. It is also important to demonstrate research fit with faculty.

GPA and grades are probably used as a filter to some degree, but my impression is that they don’t matter that much once you’re past a certain threshold (there are exceptions; I believe math PhD programs put more weight on grades and graduate coursework compared to other fields). Teaching experience might be a plus—for instance, the Princeton CS application had us write two short essays about our previous teaching experience—but it does not matter nearly as much as research experience. Some professors noted my TA experience because they considered that particular class to be complex and difficult, but it definitely felt like more of a cherry-on-the-top sort of thing. Nobody talked to me about my extracurricular activities, save for a single student interviewer who wanted to ask more about my outreach activities.


Choosing Programs to Apply to

What type of programs should you apply to?

(This subsection is specific to applicants interested in interdisciplinary research.) When I first started thinking about applying to PhD programs, I was confused on what type of programs I should be looking at. Based on my research interests, there were many different types of programs I could have found myself applying to—computer science, bioinformatics/computational biology, or genetics/genomics programs. In the end, I applied to programs across the entire spectrum, primarily based on where advisors of interest were situated.

I often found that faculty of interest would be available through multiple programs—for example, the professors I was primarily interested in working with at MIT accepted students from both the Electrical Engineering and Computer Science (EECS) department and the Computational and Systems Biology (CSB) program. There was even one professor who worked with students from five different programs (of which I applied to three): MIT EECS, MIT CSB, Harvard Bioinformatics and Integrative Genomics, Harvard-MIT Health Sciences and Technology, and MIT Mathematics. However, note that while some schools allow you to apply to multiple programs in the same year, others will restrict you to only one application.

There are certainly differences between these types of programs, even if the research you do (the most important factor) might be the same. Of the programs I applied to, the biosciences or computational biology programs have formal rotation requirements and smaller cohorts, while the computer science programs have direct admit or informal rotation systems. There are sometimes also differences in the coursework requirements, backgrounds of peers that will be in your cohort, and types of jobs available to you after graduation.

I asked many professors about the implications about choosing a particular type of program. Some felt that I should choose a biosciences program because I would build a better foundation of the underlying biology, while others said the CS degree would be better because I would receive more quantitative training and have more flexibility. Many professors said that it did not matter—the only thing that mattered was the research, so I should choose whichever department I felt more at home at. A couple of professors put it to me as, “Do you want to be a computer scientist working with genomic data, or do you want to be geneticist doing computational work?”

One thing to note is that my academic background was entirely focused on computer science and math. I did not take a college-level biology course until the semester after I applied to graduate school, which made me concerned about whether I had a chance at the genome sciences programs. However, it turned out to be a non-issue for getting interviews. From what I was told by professors, my related research experience made up for the lack of biology coursework. So don’t be discouraged from applying if you’re in a similar situation.

How many and which schools should you apply to?

When it comes to choosing specific schools to apply to, remember that you’re primarily looking for a place with advisors you’d be excited to work with. You can get a sense of this by looking through professors’ websites and most recent publications. I also found it very helpful to go over my list with my advisors—they gave suggestions on professors I should consider working with, and sometimes advice on which specific programs I should apply to within a school.

One other aspect of choosing programs to apply to is determining how competitive your profile is. At least for me, it was difficult to objectively gauge how competitive my profile was. So once again, I would recommend talking to your advisors. They most likely have been involved in the admissions process before and have an idea of which programs would be reasonable to apply to.

Once you have an idea of some programs you’re interested in, the actual number of programs you should apply to is pretty subjective. One of my advisors gave me a list of six suggested programs and said that anything more was unneccessary. Another advisor told me ten was on the lower end—he knew someone who had applied to 50 programs (though to me that is far over the upper bound of what is reasonable).

Since this is so subjective, I only have two guidelines when choosing how many programs to apply to:

  1. Only apply to programs you would actually attend. It might seem silly I’m saying that, but I do know of people who got accepted to programs low on their list, rejected the offers because they never actually wanted to go to those programs, then tried again in the next cycle. Be sure to keep an open mind, but also be honest with yourself.
  2. Make sure your recommenders are okay with the quantity of programs—it takes time for them to edit and upload each letter, and some programs also ask them to fill out a form.

Keep in mind that if you apply to a lot of programs, it not only is a large time commitment, but also can get pretty expensive. However, programs usually offer fee waivers to those with financial need. You can also sometimes get a waiver if you participate in certain programs or attend an admissions webinar.


Statement of Purpose

By application time, the statement of purpose is the last major part of the application you have control over (except for perhaps contacting professors). Everything else—transcripts, letters of recommendation, your CV, publications and presentations—is more or less determined by your work from the past years. For me, that made the statement very daunting to approach, but you essentially are conveying what you want to research, why you want to pursue a PhD/this research area at that institution, and how your past experiences make you qualified to do so.

There are many, many online resources with advice on how to write the statement, but these resources were particularly useful to me:

Here are my own tips for the statement of purpose (with the disclaimer that this is just my opinion, and there are differing perspectives that are also successful):

  • Don’t put off writing the statement for too long. The first draft will be bad. That’s okay, just get it out of the way so you can revise it into something better. At the very least, make sure you start early enough to have time for multiple revisions and to get feedback from current graduate students and from professors. I asked my friends and current graduate students to help with the earlier versions and sent the final versions to my professors.

  • The most helpful advice I received was from one of my advisors saying that your essay needs to quickly tell them what is strongest about your application. I originally opened my statement with why I found computational genomics fascinating, but my advisor commented that the committees are reading hundreds of applications, and you cannot afford to bore them with the same language and content that they’ve seen a hundred times. Think of the first paragraph as an abstract that shows clear signals of what makes your candidacy special. For applicants in CS, I’ve also heard it’s useful to indicate your subfield in the opening paragraph so that the committee knows which faculty they should route your application to.

  • When describing your relevant previous projects, try to make it clear and understandable for someone not necessarily in that specific area. What were the questions you were trying to answer, and what were the broader impacts of your project? How did you demonstrate independence? For my completely unrelated research experience, I tried to instead describe those projects very briefly and then emphasize the transferable skills.

  • Some schools don’t have any length limits, but still keep it reasonable (I aimed for two single-spaced pages at size 12 font). Also, the occasional school (e.g., MIT) might have you enter the text of your essay into a plaintext box on the application, so if you add formatting to your essay, make sure the statement can still stand without it.

I have uploaded one version of my own statement here as an example, but looking through this database of other SOPs will quickly reveal that there are a lot of different successful SOP writing styles. My own statement is certainly far from perfect—I think that if I had to go through the application process again, I would focus more on future research topics of interest and what I hope to contribute.


Letters of Recommendation

Ask early (one month ahead at minimum; somewhere closer to two or three months ahead is better), send an organized spreadsheet of deadlines, and thank your letter writers afterward. The best types of letters are from professors that worked with you on research and know you well. If possible, avoid gathering letters from professors that only know you from class (math programs are sometimes an exception, where higher emphasis is placed on strong performance in difficult graduate classes). It does help if your recommenders are familiar to the admissions committees, but the priority is still to find recommenders who know you well.

At one of the interview weekends I attended, the director told us that because there was so much inflation in the quality of applicants across all areas (grades, research experience, etc.), you were only invited to interview if one of your letter writers essentially said you were the best student they had ever worked with. This program was very competitive, so this is perhaps an extreme example, and such a letter is not required to get accepted at every program. But the point stands: these letters and what they say about your ability to do research are important.

In my case, my first letter was from my primary advisor who I had worked very closely with in the relevant research area for years. That letter probably carried the most weight. My second letter was from a professor in a completely different scientific area (physical oceanography). However, he had worked with me very closely and was able to write about research skills that would transfer to computational biology. I did not conduct research with my last letter writer, but she knew me very well. I had been both a student in and a TA for her class, and she also ran the volunteering and outreach activities I participated in.

I thought about replacing one of the latter letters with one from another professor who had worked with me in the relevant research area. But I ultimately chose this combination because these professors knew me the best. Since I’ve never been on an admissions committee, I honestly don’t know how much this mattered or if I should’ve chosen a different combination. But you can see other people’s advice on choosing letter writers in the guides linked at the end of this post.


Standardized Tests

I did not take any standardized tests. There’s a very good chance you also won’t have to if you already meet each program’s English proficiency requirements. Of the 16 programs I applied to, not a single one required the GRE (and many said not to send a score because they won’t even look at it). However, make sure you double check the requirements of every program you’re interested in. For instance, in the year I applied, some of the CMU School of Computer Science programs “highly” or “strongly” recommended GRE scores and asked you to provide a reason if you did not send a score.

If you are an international student, you might have to take the TOEFL or IELTS to meet English proficiency requirements.


Contacting Professors

Before you email them, make sure the professor is receptive to cold emails. There’s often a section on their contact page related to potential students that tells you whether you can email them and what you should send them. For guidelines on what to write in the email, read this article by a UVA professor. Don’t be discouraged from applying if the professors don’t respond. At some of the programs I was accepted to, the professors responded to my email only after acceptance (or never responded at all).

For my own applications, I didn’t end up emailing many professors. In particular, all of the biosciences programs I applied to did admissions through a central committee, meaning that emailing individual professors would not have affected admissions (but you can still talk to them about upcoming research projects, whether they will be taking students, etc.). However, there were some CS programs where admissions were done at the advisor level—you were only accepted if there was at least one professor who chose your application and wanted to advise you. In that case, emailing professors can sometimes be important, but this still varies by program. You can sometimes find out if you should contact professors by looking at the professors’ contact pages, the program’s FAQ section on their website, or attending an admissions webinar.


Waiting it Out

The biosciences deadlines mostly fell around December 1, while the computer science deadlines tended to be on December 15. I remember looking forward to submitting my last application, thinking that I would finally be able to relax. Instead, the waiting period anxiety ended up being the worst part of the process for me. If you’re similarly prone to stressing out over this, I would come up with a list of things you’ll do to distract yourself until you hear back, like focusing on hobbies or applying to summer positions.

There were two grad school admissions forums I was aware of when applying, and I have mixed feelings toward them. On one hand, I absorbed a lot of helpful advice that I applied to my applications. On the other, these forums are notorious for feeding into the frenzy and anxiety as people start hearing back from programs—especially when users make up results to mess with other people. As such, I feel an obligation to not mention these forums by name.

But in case you’ve already stumbled upon them: for one of the forums, I enabled email notifications for the computational biology thread, and it actually really helped my anxiety because I stopped checking unless I got an email. I did check the other forum way too frequently, but to me, hearing that other people got invitations to programs and I didn’t was better than wondering for weeks if I would hear anything. I think this kind of thing varies by person, so my only advice is to make a plan for the waiting period before you finish applying and stick to it (e.g., say that you’ll only check the forum once a day in the evening).


Interview and Visit Weekends

Primarily, these visits are an an important opportunity to chat with advisors and students to learn what life would be like at that program, so be sure to ask questions (you can find some examples at the bottom of this list).

The convention and timeline for visits was different between bioscience programs and CS programs, so I’m going to split them into separate sections. But in both cases, the visits are a ton of fun—besides meeting your potential advisors, you get to meet and hang out with your prospective cohort and current graduate students, and you’ll get a feeling for what it’s like in that city.

Biosciences Programs: Interview Weekends

Almost every single genome science or computational biology program I applied to held formal interviews at the department level (the exception was the joint CMU-Pitt computational biology program, which interviews through Pitt but admits students without interviews through the CMU side). I received interview invitations from mid-December until mid-January, and the interviews happened between early January and the end of February. The schools would fly us in for a few days, where the events usually consisted of a welcome presentation by the directors, the actual interviews, research presentations by faculty/students, a student panel, events in the city with current graduate students, and a fancy dinner. For the programs with online interviews, they invited admitted students to a later in-person visit.

The post-interview acceptance rate varies by program. At one of the programs I visited, we were told by graduate students that we were essentially in at that point and it was a formality. At another program, the post-interview acceptance rate was about 33%. I mention this because people often wonder after receiving an interview invitation, but try not to think about this and just focus on preparing for the interviews.

The interviews themselves were 30-minute conversations with 3 to 6 faculty (you usually have some say in which professors you’ll talk to), and a couple programs also had a student interviewer. I found this practice list of questions on Reddit to be very helpful when preparing (though it’s definitely overkill; I wasn’t asked most of the questions on that list).

Here are questions I was asked at almost every program that you need to be prepared to answer:

  • Tell me about a research project that you worked on.
  • Why do you want to do a PhD?
  • What are your future career plans?
  • What are your future research interests, and which faculty are you interested in working with?
  • Do you have any questions for me?

Some questions I was asked pretty frequently:

  • What was one challenge you experienced during your research project?
  • How did you gain an interest in computational biology/genomics?
  • Why this program?

Some questions that were uncommon but might require preparation:

  • Tell me about a research paper you recently read.
  • Tell me about a research paper you recently read outside of your research area.
  • What research papers have you read from my group?
  • Propose two projects that you would love to work on in your PhD: one related to [your past research area], and one in an unrelated area.

The biggest tip I have for these interviews is to read some papers from your research groups of interest beforehand, especially if you’ll be interviewed by those professors. One of my advisors told me that most applicants fail to do so, and so it can make you stand out if you do. The professors I talked to explained their research as if they didn’t expect me to know anything, but it makes a very good impression if you can ask insightful questions about past/future work.

The other big thing I want to impart here is don’t be intimidated! I know that’s much easier said than done—the nights before the interviews at my top choices, I was so nervous that I wasn’t able to sleep. But as long as you can articulate and answer basic questions about your past research and future research interests, there isn’t anything to be scared of.

For the US programs, the vast majority of interviews felt mostly like casual conversations (e.g., after discussing research, a lot of professors started talking to me about the weather because it was freezing outside and I’m from Austin). I was asked many questions about my research, but it genuinely felt like the interviewers were trying to understand my work rather than quiz me. If you have virtual interviews, I found it helpful to have slides from an old presentation ready, but it is certainly not expected.

I did also apply to two programs in the UK, and I found those interviews to be a bit more intense. The interview process also included panel interviews and technical questions unrelated to my research (the type of interview where they ask progressively harder questions until you say, “I don’t know”). However, most of the discussion in these interviews was similarly focused on my past research and future research interests.

Good luck with your interviews! Even if the final outcome isn’t what you’d hoped for, remember that there are a lot of factors outside of your control that don’t have anything to do with your own abilities, such as whether your advisors of interest are seeking students this year. At one of the interview weekends I attended, the program director emphasized how random the selection process at this stage can be. He said that when the admissions committee members finalize their candidate rankings, the lists are often completely different from each other, so don’t take the final result too personally—if you made it to the interviews, you are already qualified and have what it takes.

Computer Science Programs: Visit Weekends

Unlike the biosciences programs, the computer science programs did not hold interview weekends. Instead, from early January to early February, I received invitations from individual professors to speak for 15 to 30 minutes over Zoom—sometimes on behalf of all the computational biology faculty, and sometimes on behalf of their own research group. The interviews themselves were all pretty much the same questions: tell me about a research project you’ve worked on before, why do you want to get a PhD, which professors are you interested in working with, and so on. One of the meetings took me by surprise because it was basically just the professors telling me about the program and asking if I had any questions for them—so come prepared with questions.

After being admitted to a school, I received information about visit days, which occurred in February and March. The visit days usually consisted of a welcome presentation, many one-on-one meetings with faculty and current students/postdocs, a fancy dinner, events with students, campus tours, a student panel, and presentations about the program (e.g., student groups you can join, degree requirements, housing).

Make sure you ask all the questions you need to make your grad school decision. It’s an important decision, so I did feel a bit stressed out at some of these visits. But also enjoy yourself! The programs will usually organize fun events for you to get to know prospective and current students. For instance, one school in California planned a bonfire at the beach and an evening at the aquarium, and another school held a party at their school museum.


Waitlists

I was waitlisted at one school in February. I didn’t hear from them again until the evening of April 14 (which was the day before the deadline for us to respond to funded offers), when they called to say they were accepting me—so don’t lose hope until you actually receive a rejection! In my case, I had assumed I was rejected when I didn’t hear anything for so long, which meant I had already accepted another school’s offer.

This situation was my fault for not communicating with the program throughout the waiting period. If you are in a similar position, maintain communication with the program so you can know your status and inform them if you are no longer interested. Be prepared to also make your decision with short notice, and to already be thinking about whether you would accept the offer.

For people who are waitlisted at their top choice: at one of the interview weekends I attended, the director told us that if you’re waitlisted, it can help if you tell them you will certainly accept an offer.


Deciding Between Programs

Technically, you have until April 15 to make your decision. However, I would highly recommend declining an offer as soon as you know you aren’t going. Sometimes, declining an offer means the program can accept people off the waitlist, but if you decline too late then the program might not be able to offer them your spot. So it really can only help other people to decline offers you are not interested in as soon as possible (that being said, don’t let anyone pressure you into rushing your decision).

When choosing a program, it’s almost always advised that the most important factor is your potential advisors (assuming all programs will adequately fund you throughout your PhD). You should make sure there are multiple advisors you would be happy to work with at the program, taking into account not just research fit and the type of problems you would be working on, but also factors like advising style, lab culture, alumni outcomes, and the advisor’s expectations for grad students (which you hopefully were able to glean from visit days and talking to current grad students). That being said, things like location, department culture, and program requirements are important as well. Your current advisors can also provide a well-informed perspective on your options.

If you’re unsure of where you should go, I like the burning spreadsheet approach: fill out information about the different factors for each program, maybe even assign rankings and weights for each category to calculate an overall score, then burn/disregard the spreadsheet. The idea is that this lets you absorb all of the important information and understand what is important to you, but at the end of the day it’s not a 100% logical and rational decision that can be broken down into numbers. If you start fiddling with the weights to change the outcome of your spreadsheet, then you’ve probably already made your decision.

For me, I was ultimately torn between three programs for a long time, since they were all fantastic but had different strengths. Two were even in the same city and had overlap in the professors I was allowed to be advised by, making the decision even harder. I kept going back and forth, agonizing over the pros and cons of each program and how much weight to give each factor. People were constantly telling me that I couldn’t go wrong with any of the programs, but that didn’t help me decide, since “they are all great choices” is not the same as “there is no best choice”. What helped me most was realizing that I was trying to reach something that did not exist—there was no best choice. Once I understood that, I stopped stressing out over whether I was making the wrong decision and instead went with what felt right to me and offered the most freedom/flexibility.


Miscellaneous

Remaining thoughts:

  • The interview/visit period felt pretty overwhelming for me. If you also end up applying to many programs while still in school, I would recommend making your last semester course load as light as possible. It’s not just a matter of missing class and taking time to prep for each interview, but also that it’s really hard to mentally focus on schoolwork with multiple interviews and trips.

  • Spinning the unusual parts of your background in a positive light can help you stand out. For example, I have an English degree and was slightly worried it would end up hurting me, since it meant I took a lot of unrelated courses in my later years. However, some of the professors I talked to were very excited about it, telling me that it can be hard to find people with both technical and writing skills.

  • If you haven’t put together a CV yet, you can browse the CV templates on Overleaf. I really liked and modified the template by Charles Rambo.

  • It took longer than I expected to actually fill out each school’s application form, so start with ample time before the deadlines.

  • I was very fortunate to have had the guidance of many amazing professors and graduate students during my own application cycle. If you have any lingering questions about the application process or grad school, feel free to contact me.


Other Resources for CS Applicants

A list of guides and articles I read when applying:

Application assistance programs are run by graduate students and allow you to get feedback and guidance on your application from an assigned mentor. Here are examples of some programs (check their websites early in the application cycle for registration):


April 16, 2023