AnonNeuroGrad

Sure I just wanted to make sure you didn't only apply to top schools having heard the advice that "GPA doesn't matter if you have strong research" which I made the mistake of. If you're in NIH PREP/IRTA that's a big plus. Just make sure to appeal to at least 3 professors at each school you really want: I had a few PI's want me (and tell me as much) but because they were only 1 or 2 per school, I never ended up getting in (*cough* Carnegie Mellon *cough*); mea culpa. I have some specific recommendations regarding Columbia if you want to PM me and I can help you through which PI's to talk to or how BU works if you want as well. Best of luck!!!

It depends on your department and I'm not sure about Pitt CNUP. In CMU Biological Sciences, you have to rotate twice first within their department and then can look to ones outside possibly even at Pitt. I've heard rumored that it can be harder to join a CMU lab from Pitt but I've also heard to opposite. You can become affiliated with the CNBC after your first year I think after filling out a non-competitive application. All the CNBC professors you can work with even if you're not CNBC but that again goes back to how your department works. Of course, they only regulate your primary advisor that's not to say you can't pick up a secondary one or just collaborate anyways. I think the CNBC is good in that they have some institutional support (maybe small fellowships or travel grants) they can offer and regularly meet up in Mellon. Also know that the CNBC has it's own (highly-selective) PhD program called the Program in Neural Computation that has a separate application from either CMU or Pitt but should you be admitted, you would be a student of both institutions.

I think you are competitive for a place like BU international student status aside. It really comes down to who is writing your letters: if they're big names, you'll get interviewed; and there are some very big names at Chicago. I'm here and we also had some interview at several other of those schools as well but your stats don't see out-of-line with the G1's I know. BU has a very bimodal (?) set of faculty that they are trying to remedy in that, around the 2000's they were real hard-hitters and top in the theoretical field with people like Eric Swartz, Howard Eichenbaum, and Stephen Grossberg but the former two have now passed away and they lost some of the mid-career faculty to other departments (Barbara Shinn-Cunningham, Tim Gardner, few others). Recently they've hired a bit of rising star set of new faculty though: Ben Scott from Brody and Tank lab doing freely behaving wide field in rats; Chandramouli Chandrasekaran from Shenoy lab doing monkey prefrontal and decision-making with neuropixels (soon); Steve Ramirez from Tonegawa lab who does memory and engram formation/deletion and just won a prestigious NIH Director's Transformative Research grant; Mark Howe from Graybiel lab who does basal ganglia and decision-making/action selection in mice; and Laura Lewis who does sleep and attentional states in fMRI. They also have of course computational folk like Nancy Kopell (legend) and also Mark Kramer and Uri Eden and the math/stats department is hiring 2 more and I think are looking for comp Neuro. The psych and brain sciences is also hiring another systems neuroscientist as well. As far as who you should email, I think they're all receptive but know for certain that Ramirez and Chandrasekaran have the funding to take on new students in the next year. If you're aiming high, let me give you some suggestions of schools that are not as highly ranked and yet still very computational. University of Pittsburgh Neuroscience or Carnegie Mellon Biological Sciences: neither of these is explicitly computational or as competitive as their CNBC PNC but allow you access to the CNBC as an affiliate never the less. CMU Bio is moving into neuroscience with the starting of a new department (but not new degree program) headed by Barbara Shinn-Cunningham. GeorgiaTech/Emory: These schools are very closely affiliated and are somewhat easier to get into. They boast some good computational people like Eva Dyer, Chethan Padarinath, and Chris Rozell. Brown: everyone thinks of this as an Ivy but that label doesn't really matter in terms of grad school. I had a lot of friends get into places you listed and turn down Brown. Either way, they still have a good computational unit there I know of Serre and Sheinberg. They also have some neuroprosthetics there. USC: A relatively easier school to get into than you listed but they do have a good focus on neuroengineering there. Stony Brook: A smaller department but maybe worth applying to as they have some computational people like Giancarlo La Camera, Braden Brinkman, and Alfredo Fontanini University of Oregon: A department I think everyone should be considering which is, imo, the fastest riser in terms of the up-and-comers. Unfortunately they lost Yashar Ahmadian to Cambridge but they still have Luca Mazzucato and I expect them to fill in another computational neuroscientist. For what it's worth, they have a ton of cash from the Knight family and are doubling in size plus have all the cool tools even a lot of R1's don't have like Neuropixels and a mesoscope. Their specialty is systems Neuro but I can say that their director, David McCormick, and professors like Cris Niell push for and value computational approaches. UCSB: A stunning location and they've been growing super quickly. They were able to poach Spencer and Ikuko Smith from UNC-CH and are aggressively hiring more faculty including I think some computational. It's a direct admit program but still worth a look.

Trying not to be that guy but I'm here to tell you it's "not fine" if you have that GPA and anyone who tells you otherwise has no clue what they are talking about. That length of time will not counter your GPA and your GRE scores are not quite strong enough to offset it either especially considering that many schools are no longer using it. Neuroscience is insanely competitive right now with many/most R1's having admit rates below 5% with many being below 2%. I know people at the schools you listed and got into comparable ones but I had to do a ton to compensate for my similarly low GPA and still I didn't do well. I did end up at Boston University (a good school) but only because I networked with the right PI who pulled me into a direct-admit program because he looked past my GPA. My stats are posted in a block below. The key for you is to network and to get several PI's at schools to fall in love with you so they can pull for you no matter your GPA. I don't believe your GPA is reflective of your abilities and schools know that but the unfortunate truth is that it's hard to look past it when so many other students have higher ones (and no PI wants to knock someone for not having a "hard major") plus incoming class GPAs matter for rankings and when applying for institutional grants like the T32. Furthermore, a low GPA means that you are unlikely to get competitive individual fellowships like the GRFP, NDSEG, CSGFm Ford, and possibly the F31 so the school is less inclined to take you. I wish you all the best and feel free to message me if you have any questions. I know first-hand how difficult it is to offset a low GPA but I've been somewhat successful so hopefully I can help you out. Undergrad Institution: Top 100 liberal arts universityMajor(s): Math and BiochemistryMinor(s): NeuroscienceGPA in Major: 3.1Overall GPA: 3.2Position in Class: AverageType of Student: Domestic MaleGRE Scores (revised/old version):Q: 164 (88th percentile)V: 167 (98th percentile)W: 6.0 (99th percentile) Graduate Institution: Top 5 Public UniversityDegree: M.S. Applied MathematicsGPA: 3.7Position in Class: ?Type of Student: Domestic MaleResearch Experience: 3 academic years in neuroendocrinology and biophysics (one third-authorship in a small journal). 2 REU's. 1 2nd-authored poster at SfN.Awards/Honors/Recognitions: Several awards for travel, summer research grant, academic scholarship (half of tuition), and awards for tutoring/mentorship.Pertinent Activities or Jobs: 3.5 years of research at a premier neuroscience research institution. 2-3 posters authored (2 large-authorship and 1 first-authored). 3 publications (all large-authorship) in eLife, eNeuro, Nature with three more in preparation (1 3rd-author, 1 large-authorship, 1 potential first-authorship MS thesis) but probably no more than 1 submitted by application time. Possibly one or two more I'll be included on. Also work as a mentor for underprivileged science students. Several years of serious programming (software development-ish level) in Python with some Matlab.Special Bonus Points: Completed MS in Applied Math consisting of 36 credits (see above); very well-connected through my current institution Any Other Info That Shows Up On Your App and Might Matter: Institution is very very well-known in neuroscience and pretty much universally-lauded (you could probably guess it). Research Interests: Computational neuroscience especially dimensionality reductive approaches to large-scale datasets in mouse with calcium imaging and vision

You're aiming too high and I think would need a few years off working as a research associate if those are the schools you want to go to. Typically, those schools will see students with above a 3.5 GPA, 3 years of research, and a 160 quant score. Especially aiming for psych programs you're gonna see the average student with a 3.75. I interviewed at CMU nad have friends at UVA, Pitt, and JHU for neuro and have around a 3.8 with several years of research experience with good scores. The one (at UVA) that doesn't have a high GPA (a 3.2) has a masters. I think you should do the same. A first-authorship would help a lot.

So I actually go to BU and the vibe I got/have is not so much that the departments are competitive but that they just have different philosophies regarding training. Maybe that results in disagreements but I've only been here a week so I probably am not privy to them yet. Again, PM if you want more specific details but broadly, GPN: This is the main neuro program available and is very heavy into systems neuroscience but seems to be the best funded and supported among the three. It is housed on the main campus with most faculty in either CILSE (the brand new beautiful tower for neuro and some bio) and LSE (life sciences building next door) with some at the medical center (30-45 min shuttle ride away). They seem fairly well-organized and put on programs/events for their 10-15 students a year. Their curriculum is highly programmed which is both good and bad: you get to take a lot of classes with your cohort but you have to take a lot of the classes they want you to take. This program also has access to the computational specialization but doing so will give you only two rotations instead of three (they rotate longer). Admissions for this department is very competitive. BBC: This program is run through the Psychological and Brain Sciences department and sort of bunches you up with the developmental and clinical psych students. This program is *much* smaller than I think the other two: there are four people in my cohort and they admitted eight last year. Don't think of it so much as a cohort program but as more of a European-style process because you do not rotate and are directly admitted to the lab. Stipend I think is just short of GPN but maybe only a thousand less. There is far less programming put on by the department so I can only recommend this for the student that is able to make friends easily since no one is doing it for you. You also need to have a compelling fit with a PI to get admitted. I know about 10 students interviewed (no idea how many applicants) and 4 ended up joining. This program has a *very* flexible curriculum as only 2 classes are required (Stats for Psych) and the rest is up to you and your PI. I recommend this program for someone really who knows what to do and this comes with the benefit that you get to start researching and publishing earlier without a lot of program requirements. They have grown the department a lot lately with four newish hires (Ben Scott, Marc Howe, Chand Chandrasekaran, and Steve Ramirez) and are hiring another PI. I would also caution that this department has a traditional qualifying exam where you do a literature review and have to answer questions based on those each given to you over a few days (a lot of programs now are just taking an NRSA proposal plus thesis proposal). Not sure what the qualifying exam process is for GPN or A&N. A&N: This program is much more heavily focused on, eponymously, neuroanatomy and clinical applications. It's based on the medical campus which might be a negative for some but they do seem to have a decent amount of programming (not as much as GPN but more than BBC) for their students and also seem to have far more masters students. I think they emphasize teaching more and their students run the gross anatomy labs for the med students. A lot (all?) of MD/PhD students are through this program. This one is the most foreign to me so I have no idea what the stipend, qualifying exam, or curriculum is like here but it should be easy enough to find. Cohorts are about medium sized and are 4-8 not including a similar number of masters students. Let me know if you have any questions about these as I can relay them directly to a person in the program. Personally, I think the GPN is the best program for the average student but I liked the independence of the BBC (but really don't like that I essentially don't have a cohort but this isn't the biggest issue because I'm pretty gregarious and have a lot of Boston friends). For whichever professor you'd like to work with, just make sure they are affiliated with the right departments. If you have a lot of specific appeal with one professor, consider the BBC because all they require is for said professor to want you and have the funding available (plus a back-up professor). The GPN you need to appeal to the entire department and need to have three professors step forward and say they want to take you.

It's because you're a trained physician which I have no idea what to make of especially with you being from India. Also, the cognition of music is a very odd (not bad just uncommon) thing to be interested in and might restrict your options.

I'm in BU BBC and I know a fair bit about all three programs (my lab is in A&N and I interviewed with GPN [but didn't get in') but you'll have to PM me if you want info and I'll give you my personal email.

So your profile is a bit odd to me and is very atypical for a neuroscience applicant in the U.S. If I was a committee member, I'd like to see specific research experience in a traditional neuroscience lab; was this your experience? What journals are you authored in and are they regional ones from India? I know of someone with a very similar and profile as you (might even be you) and they are getting a master's in Europe to allay such concerns. I also don't know if an MBBS from India is of any value and committee's are a conservative bunch when it comes to addressing degrees they don't know about. Also, the schools you've picked are sort of all over the place and very top-heavy from those in the top-10 (Harvard, MIT, Princeton, and Columbia) to those in the top-20 (Chicago, Northwestern) to those in the top-50 (Rochester, and Cornell) and then some top-100 (Tufts and UConn) to Nevada whose program I've never heard of. I think you need to try to apply to more safety school's in the top-100 because It's already super hard to get in as an international student and doubly so because you don't have a Western degree.

Well what are you trying to do long-term? An MS in neuroscience is largely unhelpful especially given that you already have an MS in psych which is a neighboring field. If you have to pay for the MS in neuro, I'd say don't bother. Even if you didn't have to pay, I'd probably also say don't bother. Really the only thing you should be going into now is either working in industry or a PhD program; to that end, I'd advise you just to try to enter a neuroscience lab through your psych PhD.

Your GRE doesn't generally matter although tour quant is pretty low; it won't get you disqualified if your other stats are strong but it will be a reason to deny you if they're deciding between you and another candidate. Going for top-tier places like Princeton or Duke, it will really depend on your other stats.

Nothing related to neuroscience probably unless you're a scientist at one of the few companies that do it (someplace like Inscopix or NeuroPhotometrics or a pharma) or a staff scientist at one of the places like the Allen Institute. Most everyone just leaves is what I hear.

Good luck to you all! So many memories, so many tears; I miss it in some sick way.

Straight to PhD. MS's cost money usually and you're getting equivalent experience in your neuro postbacc. Really the MS is only for those who have zero science experience and are looking to switch into neuro.

A few of those programs are harder than you think to get into and I would probably move them into Reach programs namely NYU and Chicago possibly also BU (I might be biased but my interview cohort there were also interviewing at places like NYU, Northwestern, MIT, etc.). I think it doesn't have as prestigious of a reputation but it draws some of the same caliber of applicants as those applying to the two big name Boston schools (MIT and Harvard) who use it as their "back-up"; a lot of the people I met decided to go to places like UCSD, UCSF, Yale, Karolinska, etc but nevertheless you'd have to compete against them for a spot at the interview. We had >500 applicants, 40 interviewees (2 international), and 9 were in the cohort (2, like me, were swung into a different neuro program). USC should not be reach imo. If you're looking at computational neuro, you should remove WSU and Yale. You didn't ask but here's a list of schools I made for myself when applying. There's a lot of systems neuro mixed in here. I hate lists because it is so much more nuanced than that. I focused my applications based on particular papers and research collaborations that stood out to me (which is what you should do as well). I'm about to tell you a list (in no particular order) of places that do great work in comp. neuro. but this shouldn't be taken as comprehensive and know that I am biased by groups doing systems work as well in what's called "latent space inference" or "manifold learning" in the United States. Look at who is presenting at Cosyne/CNS/CCN for a good idea of comp. neuro. I don't really make a distinction between many computational and systems neuro since they're so tightly integrated. Adam Calhoun has a great blog post on this and the network graph shows you the current who's whom of the field and their main collaborators. I've starred the "top five schools" imo. I'm intentionally making a long-list so that you can decide who are the top ones in your desired subfield. If I didn't give PIs, it's because I'm lazy. Here's my top ten list Stanford* (Surya Ganguli, Krishna Shenoy, Dan Yamins, Scott Linderman, EJ Chichilnisky, Google guys like David Sussillo and Jon Shlens) Princeton* (Jonathan Pillow, William Bialek, Carlos Brody, David Tank, Sebastian Seung) NYU (Wei Ji Ma, Xiao-Jing Wang, Dmitri Chklovskii, Eero Simoncelli, Dora Angelaki) Columbia* (John P. Cunningham, Mark Churchland, Liam Paninski, Stefano Fusi, Ken Miller, Randy Bruno, Niko Kriegskorte, Larry Abbott) MIT (Mehrdad Jazayeri, James DiCarlo, Joshua Tenenbaum, Michale Fee, Mriganka Sur, Mark Bear, Haim Sompolinsky [half-time in Haifa, Israel]) Washington University in St. Louis (James Fitzpatrick, Adam Kepecs, Ilya Monosov, Camillo Padoa-Schioppa, Lawrence Snyder, Geoffrey Goodhill, Keith Hengen) University of Washington (Adrienne Fairhall, Rajesh Rao, Eric Shea-Brown, [many Allen Institute scientists]) UC Berkeley (Hillel Adesnik, Jack Gallant, Bruno Olshausen, Fritz Sommer, Linda Wilbrecht) Carnegie Mellon/Pittsburgh* (CNBC) (Brent Doiron, Rob Kass, Bard Ermentrout, Byron Yu) University of Chicago (John Maunsell, Jack Cowan, David Freedman) And an extended list including many systems people UC San Diego (Tatyana Sharpee, Terry Sejnowski, Thomas Albright, Bradley Voytek) Harvard University (Sam Gershman, Chris Harvey, Cengiz Pehlevan, Mackenzie Mathis, many others) Caltech (Markus Meister, Doris Tsao) Champalimaud Centre for the Unknown (Christian Machens, Zach Mainen, Leopoldo Petreanu) UCLA (Jonathan Kao, Dario Ringach, Josh Trachtenberg) UCSF (Massimo Scanziani, Vikas Sohaal) ETH Zurich (Ben Grewe, Valerio Mante) UCL* (Gatsby Unit and Sainsbury-Wellcome) (Matteo Carandini, Ken Harris, Maneesh Sahani) Oxford (Tim Behrens, Tim Vogels, Adam Packer) HHMI/Johns Hopkins (Janelia Farm) (Marius Pachitariu, Vivek Jayaraman, Davi Bock, Karel Svoboda, Reza Shadmehr) Cold Spring Harbor Labs (shared students with SUNY Stony Brook sometimes) (Anne Churchland, Anthony Zador, Tatiana Engel, Adam Kepecs) Boston University (Chandramouli Chandrasekaran, Ben Scott, Nancy Kopell, Mark Kramer, Uri Eden) Toronto (Joel Zylberberg [affiliated through York]) Penn (Diego Contreras, Josh Gold, Konrad Kording) University of Rochester (Adam Snyder, Thomas Howard, Greg DeAngelis) Baylor (Andreas Tolias, Jacob Reimer) Northwestern (Daniel Dombeck, Andrew Miri, Sara Solla) Brandeis (Eve Marder, Gina Turrigiano) Brown (Jerome Sanes, David Sheinberg, Wilson Truccolo) UT Austin (Alex Huk, Dana Ballard) Tons of Germans that I have no idea what universities they are at EPFL An extended extended list of places that still do this work and are up-and-coming or may be more systems oriented or I know less about but are worth applying to Georgia Tech (Chethan Padarinath, Eva Dyer, Chris Rozell) University of Oregon (Cris Niell, David McCormick, Yashar Ahmadian, Luca Mazzucato) SUNY Stony Brook (shares with CSHL) (Memming Park, Giancarlo La Camera, Alfredo Fontanini, Braden Brinkman) UC Davis University of Southern California Duke UNC Chapel Hill Yale UT Houston Rice UCSB

I had a 3.1 cGPA with a 2.8 in my math major and a 2.9 in my biochem major; I got four interviews (BU, Carnegie Mellon, UO, and Stony Brook) and into BU, UO, and Stony Brook. I'll be attending BU which is just about in the top-20 nationally and one of the stronger ones for computational neuroscience (which is my interest). I will say that, first of all, you can still not only go to grad school but also a good one but it won't be easy by any means. I took four years off and I had a bit higher GPA and did 3 years of research in undergrad so, for you, probably expect to take at least that long. One of my undergrad projects turned into a 3rd author publication and then I was able to secure an RA position after college in neuroscience research. I spent that time getting two more publications and an MS over the four years; although I didn't get interviewed for most top schools I applied to, the PI's I met were super enthusiastic about having me there because of my past experience and MS. So I guess this is to say that it can still be done and hopefully neuro won't get significantly harder in a few years. PM me if you have any questions but I almost never am on here! I have Discord at the same screen-name so find me there if you have questions.

Yeah he has really been pulling for me and was fairly disappointed I didn't get admitted to the GPN; this year, BU apparently had >500 applicants but only wanted to take a class of 8; it was insane.

Thanks yeah I just found out from the Chair that there are no hard rules for this and it's really up to the individual PI (it's direct admit). I had a good conversation with the PI and they think that matching the stipend of the GPN students is perfectly reasonable.

Yeah I’m just worried about the stipend: at UW, I’ve heard their psych stipend is $8,000 lower than the neuro one. I’m also worried about the TA requirement and the different community as I’ve heard they have less programming (in the number of student activities sense) than the GPN. I know that they are also direct admit but it’s to the professor I most preferred. Just asking because I’m trying to reduce the information gap as early as possible because I’m still waitlisted for two schools and will need to make a quick decision.

Seems like there's not many people on the thread anymore but does anyone have any information comparing BU's GPN and BU's BBC programs for neuroscience? I'm still waitlisted for the GPN but have received the generous offer to be let into the BBC program if I don't get admitted to the GPN. Thanks in advance!

NYU has very strong computational work but a lot of their scientists do old school stuff and have had their time in the sun (sorta Buzsaki, Movshon, Rinzel) and I doubt Yann LeCun would have time to work with anyone. Rumor has it also that XJ Wang is leaving but I didn't tell you that. Simoncelli, Chklovskii, and Wei Ji Ma are all good options still but their research is somewhat different. It is a Swartz center though and NYC is awesome! They just hired Christine Constantinople who has some good systems ideas using automated behavior training for rats I think. UCL is obviously excellent with leaders in both theory (Sahani, Latham, Hyvärinen, Botvinick, Pouget) and systems (Harris, Carandini, Mrsic-Flogel) and I'd rate it #1 in the world for combining these approaches. Not sure how Brexit will affect things though and being so close to DeepMind means good industry connections (and also that faculty are harder to keep). I will once again plug CMU/Pitt's PNC program: they're planning on growing CMU's integration of neuroscience and machine learning through a new Institute of Neuroscience (headed by Barbara Shinn-Cunningham) while still maintaining the joint venture (the CNBC means effectively Pitt and CMU are combined and it's funded by a T32). The work of Byron Yu, Brent Doiron, Bard Ermentrout, and Robert Kass needs no introduction but I've heard (from a postdoc who interviewed in ML) that they are trying to double the size of their ML department which is already #1 in the world. The CMU side of things will also be hiring more neuroscientists for collaborative work and they just built a new business school for tech transfer.

I visited Boston and I think you should take into consideration that the city is pervasively academic what with Harvard next door and Boston University across the river not to mention Tufts, Northwestern, Boston College, and the numerous biotechs all in close proximity. I personally loved that atmosphere whereas I've heard that Pasadena is a bit out there. Given the choice, I'd pick MIT as the ability to go over to Harvard, BU, or the medical campuses is not to be discounted in terms of the ability to network.

I agree with glialstar here. I'd def pick option (2) or (3). Between getting a 4.0 in an MS or having a paper (remember, 1st author!)/getting an LoR from a big name, I'd pick the latter 10/10 times. 3 years is an acceptable time frame that you'd potentially have an accepted paper with a bit of luck and a willing advisor. I think if you can get a big name advisor with a glowing letter, that's the best option followed by the 1st author followed by the MS. I'll just reiterate that MS GPA's imo do little to overshadow a poor undergrad GPA despite what people might be saying, at least in my experience.

So I have some good information for you since I was in your exact position when I graduated (almost exact same stats) and probably would’ve had the same results. I’ll help walk you through what I did to compensate for low grades and what worked/didn’t work and how I did this application season. i graduated from a regional SLAC with a 3.2 GPA in Biochemistry and Math with a minor in neuroscience. I had three years of undergrad experience but no pubs. I went on to take the GRE and got a 167/164/6.0 I think V/Q/W. I thought deeply about where I needed to compensate and how by following the advice of others and some good blog posts; my weakest areas were going to be my GPA, lack of publications, and letters of rec (no one knows of my recommenders). Thus, number one and two on my list were to enroll in an MS program with good grades and to get some publications. Also to find a job with some big names to write my letters. I was actually fairly successful at everything: at the time of application, I was a quarter shy of finishing a masters in applied math with a 3.7 at a top-3 public school and I had been working 3.5 years at the premier neuroscience non-profit gaining 3 publications (3rd in Neurotox., mid in eLife, mid in eNeuro), 2 in review (mid in Nature, 2nd in PLOS ONE), and two more in prep (3rd and a 1st). My letter writers were all now former academics (not PIs though, postdocs) that were known in the community. I thought I did everything I could have almost as best as possible and all my mentors had told me that I should shoot very highly in terms of schools. I applied to the following 14 schools: Harvard, Stanford, Columbia, Princeton, Carnegie Mellon/Pitt, UCSF, UCSD, Carnegie Mellon (Bio), University of Oregon, Boston University, Stony Brook University, MIT, and NYU. I also networked heavily and emailed 2-3 PIs at every school and corresponded with around a dozen extensively over email, over Skype, or in-person at SfN. I was feeling very confident I’d get interviews from around half my schools. Now a few months later I was rejected from ten schools and received interviews from Carnegie Mellon (Bio), Stony Brook University, University of Oregon, and Boston University. I was admitted to both University of Oregon and Stony Brook University with fellowships about $5k a year each for three years. I’m still waiting to hear back from Carnegie Mellon (Bio) and am waitlisted for Boston University (GPN). Some takeaways, I should’ve asked my letter writers to address my grades and I should’ve talked more at length about why my grades were so low (undiagnosed sleep disorder [DSPD], and trauma with my best friend dying in a car accident). Admissions committees didn’t care that I got a 3.7 in my masters it felt like and this didn’t offset my undergrad grades as far as I could tell The admissions process is tightly controlled at a lot of schools by a single-person or a small committee and even though I was invited to interview (even being rated first in the cohort by a professor), it once again came down to grades and I suppose fit which was disappointing. I had believed that once I interviewed, the schools already wouldn’t care about grades and thought I was a good fit. as far as fit goes, I was a bit too specific and I suppose seemingly inflexible during interviews: I thought being a very good specific fit for two or three professors was the way to go but I’m thinking it’s better to have broad appeal. I wasn’t able to get un-pigeonholed as a “biology” or “computational” guy in different contexts. Some PIs expressed doubt as to whether I would even want to do experiments while others didn’t comment on my 5 years of programming or my masters in math. my publications didn’t matter. I never got a single comment on any of them and a lot of interviewees seemed like they had pulled up my CV in their computer a minute before the interview. I think the only sort of pub that matters is a 2nd or 3rd author in a high-tier journal or a 1st in a mid-tier journal. my work experience didn’t matter per se. I had a lot of friends get admitted to schools I got rejected from when they had only 2 years of experience and I had very nearly 7. The only difference between us was GPA so I think the advice “research experience is more important than grades” is false. it matters tremendously who is writing your letters. I’ve seen students with a 3.4 gpa and mediocre scores/experiences with letters from HHMI or NAS members get into every school in the top10. If I were to do it all again, I would have addressed my grades more directly and had my letter writers do so as well I would have not done the MS (or done a 1-year full-time) and then focused two years on getting one or more first-authorships. In retrospect, I had all the experimental and analytic skills to do so but I was just intimidated by the idea of it. I would have chased after working with big names in the field to get a recommendation from them. It matters more than it should. I wish I knew that everyone who was giving me advice that was last in admissions 5 years ago has outdated information. It's at least twice as competitive now and 1st-authorships are now going from unheard of to uncommon. I should've worked like my life depended on it and that's saying a lot because there were many weeks I was working 60-80 hours with work and homework combined. Neuroscience is the hardest life sciences field right now (except clinical psych PhDs) and possibly all of science save except some fields like ML/CS but, unless you talk to someone who's recently applied or is on admissions, you wouldn't know it. BU got >500 applications for 40 interview spots and 8 spots in their incoming class. I networked a ton but I should've networked even more. Really making sure these PIs were invested in having me. One PI at each Boston and CMU had told me explicitly they "golden buzzered" me into interview which I would've never gotten otherwise. Most of my schools were top-20 so I needed this sort of help for each one but even one isn't a guarantee especially if their admissions process is more committee based. BU's seemed to give more power to individual PI's while others were tightly controlled and voted on by each committee. I had a chat for 30 min in-person with the director at Harvard and he said he liked my skillset and would look for my application but alas I never got in (I should've been upfront about my grades). I should've applied to lower-ranked but still very good schools like Pitt, Northwestern, Rochester, UT Austin, Georgia Tech, etc. Very important: I should have applied to more biology programs. They're easier to get into and the only difference is in curriculum (work with all the same professors). Several neuro interviewees I was with at sort-of mid-level institutions also had interviews at top20/10 biology programs. I'm also going to disagree with the advice that getting an acceptance after interview is "yours to lose" because at several of my interviews the acceptance rate was below 50% and so it came down to fit mostly even if you were a wonderful person. One school told me they wanted to see at least three PI's throw their hat in the ring for you. At several of my interviews, all of us were confident, knowledgeable about our work, skilled, and driven no question but yet more than half of us will not receive an offer. This does vary though as I know of one top school that routinely offers every interviewee that passes the "is this person at least normal" test. Your grades aren't as bad as mines but I feel like for those with the same grades, the only way out is to have something that makes the reviewers go "holy shit" i.e. one or more glowing recs from big names in the field or having first-authorship. That's all I can think up right now but let me know if you have any questions.

I don't attend but it has one of the very top computational units for studying the mouse what with the QCN and Professors Brody, Tank, Bialek, Pillow, and Seung. Most other big places doing computational work, except UCL, have a focus on macaque i.e. CMU (Smith, Yu, Batista, Chase), Stanford (Ganguli, Shenoy, Linderman, Yamins), or Columbia (Churchland, Paninski, Miller, Fusi, Kriegskorte). They also have some faculty that are big names doing cog. neuro. like Daw, Cohen, and Niv. I was also super impressed with the work of new faculty Dr. Catharine Pena who really skillfully integrates transcriptomics with early environmental/social trauma in mice. I also need not remind you that Princeton is a very difficult place to get into so I hope your application profile is stellar.