Computational Certificate

Visual overview of Certificate components described in greater detail below

PiN students have the opportunity to earn a Certificate in Computational Neuroscience (CiCN). Computational neuroscience is a rapidly expanding and important subfield within neuroscience and trainees increasingly need more advanced quantitative, computational, and modeling skills to pursue their desired research. This certificate was designed to guide students, including newcomers to computational neuroscience and students in experimental labs, as they learn and practice these skills, and support them through mentorship and peer support.


As part of the Certificate, students complete coursework to learn programming, foundational quantitative subjects such as statistics and linear algebra, core computational neuroscience, and more tailored advanced topics. A series of workshops covers more specialized topics, including the design of neuroscience modeling projects and strategies for organizing computational work to prioritize reproducibility. Building on this foundation, students pursue a computational project as part of their dissertation research (although students do not need to pursue purely computational research or be in a computational lab). Critically, Certificate directors, Jan Drugowitsch and Sam Gershman, and faculty with computational expertise mentor students throughout this process so they receive guidance and feedback on their computational research. Students also receive feedback from the larger Harvard community through participation in yearly symposiums. Through participation in the Certificate, students gain the skill set, experience, and confidence necessary to pursue computational research.


Phase 1

This phase is more informal: it allows Certificate leadership to connect with students interested in gaining computational neuroscience skills and advise them on resources/courses/etc from an early stage of graduate school via yearly meetings. During Phase 1, students learn foundational topics such as coding, linear algebra, probability, and statistics through the course Mathematical Tools for Neuroscience (details below).


Students can apply to Phase 1 at any time: we encourage application as early as possible in their graduate school career (ideally first year). Students who enroll in Phase 1 do not necessarily need to proceed to Phase 2 but will not receive the Certificate if they do not.

Download the application form

Phase 2

This phase consists of the bulk of the certificate. Students apply to Phase 2 in the fall of their 3rd year (by December 1st). Older G-year application will be considered on a case-by-case basis. In order to enroll in Phase 2, students must have taken (or be taking) Mathematical Tools for Neuroscience or have tested out. They also must have a (tentative) plan for coursework and the computational neuroscience research component. We expect students who enroll in Phase 2 to complete all requirements and earn the Certificate.


In the J-term after application, students will participate in a series of workshops. During Phase 2, students will present yearly at a computational neuroscience symposium and meet yearly with Certificate leadership. Other coursework requirements (described below) can be completed at any point during graduate school.

Download the application form



Students are expected to learn Python as this is the official programming language of the certificate (workshops and many of the courses are Python-based), although there is no official course requirement.


Students must next take Mathematical Tools for Neuroscience (NB212), a semester-long course taught in the fall that covers linear algebra, dynamical systems, probability, statistics, and basic machine learning, all in a neuroscience context. Students may test out of this requirement if the certificate co-directors determine that they are proficient in the material covered through prior study. Students then may choose one of several foundational computational neuroscience courses. Current offerings include Computational Cognitive Neuroscience (Neuro 1401), Computational Neuroscience (Neuro 131), Neural Computation (APMTH 226), and Introduction to Computational Neuroscience (Neuro 120).


In addition to Mathematical Tools for Neuroscience and one core computational neuroscience course, students are expected to complete 2 quarters of additional advanced electives. Students will consult with the CiCN Directors to select appropriate elective courses given their individual backgrounds and training goals. See the CiCN handbook for a non-exhaustive list of possible courses.


In total, the coursework requirements of the certificate are equivalent to three half courses. This constitutes one additional half course on top of core PiN requirements.


Students attend and participate in a series of workshops.  Two core workshops will be taken during 3rd year J-term (following the application to Phase 2 in the fall). Thinking about Neuroscience Models covers what types of models exist, what types of questions you can ask with models, and the process of modeling. Structuring Code & Data is a 2-day workshop focused on strategies to organize code, data, and workflows to optimize reliability and reproducibility. To see the details of the 2021 workshops, see here.


Students are expected to attend an additional workshop Towards a computational career once during their third year or after. This workshop will feature sessions on internships, on coding interview preparation, and on various possible computational career options. 


Additionally, students take part in a yearly Computational Neuroscience Symposium. Phase 2 students present their work through talks and posters at this symposium.


Students in the certificate are expected to incorporate computational or theoretical work into their dissertation research. This computational component does not have to be the whole dissertation (students do not have to join a computational lab) but should roughly equate to at least a chapter.


To help advise on this research, one member of the DAC must be a faculty member who specializes in computation or theory (professors with an experimental lab who have a focus on computational neuroscience fulfill this criterion). Additionally, students will meet at least yearly with the CiCN co-directors.