Advice for Picking a Graduate Advisor

(This post is authored by Eleazar Eskin.)

Picking an advisor has major implications for one’s graduate career and is a decision that one makes very early in graduate school. (In some cases, even before a student starts a graduate program.) Unfortunately, these decisions must be made when a student is early in their career and has the least experience and perspective on managing academic relationships. Several issues are relevant to the choices you may face in terms of lab style. Is it best to join a big lab or a small lab?  An experienced faculty or a new faculty? An established research area or a new area?

In addition to research, many interpersonal factors are important to ensuring a productive advisor-advisee relationship that will ultimately promote your successful career. How well you get along with a faculty advisor may improve the quality of mentorship you receive. Further, how well you get along with other lab members may expand your opportunities for collaborative research and publication—during graduate school and beyond. The purpose of this blog post is to give advice to new students on which factors to consider when picking an advisor. When comparing offers of admissions from different schools, I suggest considering the following questions.

What types of papers will you write?  A Ph.D. program is focused on research, and you will spend the vast majority of time in the program doing research projects that result in publications. It is critical that you enjoy writing these papers in order to maintain stamina and avoid burnout. Each lab tends to have a specific publishing culture; some specialize in writing methods papers, and others produce more results papers. Often, high-level descriptions of a lab’s research interests do not capture the flavor of the papers. To develop a better understanding of the type of work performed by your prospective lab, first read recent papers published by its members. When reading these papers, ask yourself if you would enjoy writing these kinds of papers.  Ask yourself whether or not you find the papers exciting enough to spend the next four years producing more of them.

Take-home point: Reading the most recent 5 to 10 papers published by a prospective research lab is a worthwhile investment.

What kind of publication record can you expect?  A set of publications will be the tangible outcome of your Ph.D. career. The extent and impact these publications have—in terms of impact factor, journal name recognition, number of citations, etc.—will help define the next steps of your career.  It is very important to consider what your expected publication record will be when you finish your Ph.D. A mind-boggling mystery to me is why promising graduate students join labs where they are unlikely to publish frequently enough to be competitive at the next level.

Consider the average publication record of recent lab alumni (at the time that they graduated) in order to reasonably estimate what your publication record will be. This data can be hard to obtain, so another way to estimate this is to consider the per capita publication rate. This figure can be computed by considering the number of publications published per year times the number of lab members per publication divided by the total number of lab members.  This statistic is clearly inaccurate, because some labs are comprised of staff, students, post-docs, and undergrads who each contribute differently to projects and publish at different rates. However, computing this number is still useful to get a rough estimate of what you can expect.

Very promising students often join extremely prestigious and well-funded labs. On the surface, these labs seem like a great option because many publish multiple papers per year in high impact journals like Science and Nature.  However, if the lab has more than 30 members, the per capita publication rate is low and the chances of gaining authorship on these papers is relatively small. In some high-profile labs, many students still finish their Ph.D.’s with relatively poor publication records.

Take-home point: Examining the expected publication record will give insight on the types and frequencies of authorship you may expect to gain in a prospective lab.

Where do alumni of the lab end up?  If you have a clear goal in mind for your post-Ph.D.  career, consider joining a lab that produced alumni who have accomplished similar goals. If your objective is to be a professor, it may not be the best idea to join an established lab where no past student has remained in academia. If your goal is to end up in industry, it is likely a good fit to join a lab where many alumni have placed well in industry. If your post-grad goals are unclear, it is still useful to find out where a prospective lab’s students end up after finishing their graduate degrees. In particular, it is important to know how much mentoring the advisor provides to help students figure out their future directions and achieve their goals.

Take-home point: Exploring the post-graduate career trajectories of the lab’s past graduate students will give you an idea of the type of mentorship the prospective advisor provides.

How often will you see your advisor?  There is large variance in terms of how much total time you will spend with your advisor, and the quality of your first meeting with the prospective advisor may not predict the amount of future mentorship you can expect to receive. You may have a great initial talk with your potential advisor, but you may hear from lab members that they have a sit down meeting with their advisor only once every 3 months. Advising this infrequent may not be enough for you to develop research interests, carry projects to completion, and prepare a successful post-grad career.

If you think weekly individual meetings are necessary for your success, ask current lab members if this arrangement would be possible. Current lab members will know if, for example, their advisor allows drop-in meetings to answer your questions. In addition, you should find out how hands-on the advisor is with paper writing and how timely they are with helping push projects along. In today’s world, faculty often have tremendous amounts of pressure on their schedules due to grants, administrative obligations, teaching, and other responsibilities. Commonly, lab projects will be delayed for weeks because the advisor didn’t have time to make edits on the manuscript. However, many faculty prioritize writing and return edits very quickly.

Take-home point: Current lab members will be able to tell you how much hands-on guidance in research and publishing can be expected from your prospective advisor.

What kind of training will you get and who will train you?  In addition to completing research projects, one important aspect of a graduate program is how much you will improve as a researcher and what skills you will master (and to what extent you will master them). Ph.D. programs are inherently apprenticeship programs; much of the training you receive will come from your lab responsibilities. It is important to know how much the other students have learned, and how they learned these skills. Some labs have journal clubs and other structures for additional, more comprehensive training. In some labs, the faculty directly train graduate students in skills such as analyzing and interpreting data. In other labs, you may be primarily mentored by a post-doc or other senior member of the lab. While the latter example is not necessarily a bad option, you should know in advance what you are signing up for.

Take-home point: Current and recent lab members can tell you who taught them which specific skills, and you can take into account how the lab structure would fit your learning style.

What about new faculty?  New faculty are often the best option for a graduate student, yet opportunities to work with junior scholars are very under-appreciated. Recently-hired faculty are often a great choice, because they have an extremely exciting research program that is rapidly expanding. Students who join these labs are often involved in finishing the ongoing projects that helped the faculty land their job at the current institution. When considering the expected publication record, you can consider their own publication record as factoring into your expected publication record (if the faculty is willing to include you in their projects).

New faculty often are heavily involved in training the group and spend a tremendous amount of time with their students. Compared to senior faculty, they often have many fewer responsibilities. Finally, joining a new faculty’s lab is exciting. The life of a new faculty is an adventure with many ups and downs. Publishing the lab’s first big papers, getting the first grants, setting up the lab, recruiting the first students, and teaching the first classes are all exciting (in addition to stressful) times for a new professor. Being a student of an advisor early in their career is joining this wild—and potentially very rewarding—ride.

Take-home point: Many early students of now-senior faculty have done very well. For example, out of the first 5 students that I supervised for their Ph.D.’s, 4 are currently in faculty positions.

Are these considerations the only considerations for picking an advisor? Of course not. Other important—if not more important—factors to consider include personal connection and feeling of a “good fit”. However, we hope the points above provide some foundational insight into what should be considered when picking an advisor.

UCLA Bioinformatics: The Philosophy of the Ph.D. Program

UCLA | Bioinformatics

(This post is a collaboration between the instructors of the core courses for the UCLA Bioinformatics programs: Eleazar Eskin, Chris Lee, Wei Wang, Bogdan Pasaniuc, Jason Ernst, Sriram Sankararaman, and Jessica Li, along with the current director of the program, Yi Xing.)

Bioinformatics is an interdisciplinary field that combines different aspects of quantitative sciences, such as Computer Science, Statistics, and Mathematics, with biological sciences, such as Molecular Biology and Genetics.  Training programs in quantitative sciences and biomedical sciences have very different cultures and structures, particularly at the doctoral level.  At UCLA, we aim to combine the best of both worlds with the Interdepartmental Bioinformatics Ph.D. program.

We established our Ph.D. program in 2008, and we enroll 6 to 10 Ph.D. students each year. Over 45 faculty specializing in computational and experimental biology are associated with the Bioinformatics Ph.D. program, with active research and education programs spanning biology, mathematics, engineering, and medicine. The program encompasses the breadth of the growing Bioinformatics field by offering courses from over 12 departments.  The Bioinformatics Ph.D. is not housed in any one department but is an Interdepartmental Program (IDP) whose faculty are members of 17 UCLA departments.  The IDP is an administrative unit designed for multidisciplinary academic programs.   This unit also administers the Biomedical Informatics Ph.D. program and will administer the planned Ph.D. program in Systems Biology.

For many aspects of the UCLA Bioinformatics Ph.D. program, we draw upon different ideas from the cultures of Quantitative and Biomedical training programs.

In traditional Biomedical science Ph.D. programs, the majority of a student’s training in applied sciences takes place through mentorship in the laboratory.  Students do take some courses during their first year, but these courses mainly cover recent research in the field and are often team-taught by multiple faculty.  These courses typically require only minimal work outside of class.  During the first year of the Ph.D. program, these students focus on identifying a research lab to join by completing rotations in three labs.  Starting with their second year, students become members of their chosen lab and perform research full time.

On the other hand, in traditional quantitative science Ph.D. programs, the majority of a student’s training takes place didactically through challenging coursework.  In these programs, coursework consumes at least 50% of the student’s time during their first two years.  These intensive courses are usually taught by a single instructor (or sometimes a team of two) and require substantial homework assignments, course projects, and exams.  However, the courses lay a foundation for the technical skills that will become the basis of a student’s future research.  Students admitted to these types of programs are encouraged to join the research lab of a specific professor and start research right away.

Here we describe how we combine these two cultures with the principles and philosophy that guided the design of our Ph.D. Program.

  1. Training in Methodology Development. The UCLA Bioinformatics program is uniquely focused on preparing our students to develop novel methodologies that can contribute to important biological problems.  Students who are interested in methodology development are a great fit for our program.  Our program is able to maintain this focus, because UCLA hosts many other Ph.D. programs that can accommodate students interested in Bioinformatics but prefer a program with a different, sometimes more traditional, focus. These include the recently established Genetics and Genomics Ph.D. program, which has focuses less on methodology development and prioritizes biological discovery.

    UCLA also has a broad set of other Ph.D. programs in quantitative sciences, such as Statistics and Computer Science, which also accommodate students who are interested in Ph.D. research in Bioinformatics  but are primarily interested in a quantitative sciences training program.  UCLA also offers Ph.D.  programs in Biomathematics, Biomedical Informatics and Biostatistics for students interested in other areas of Computational Biology.  In addition, a new graduate program in Systems Biology is being developed in conjunction with the Bioinformatics IDP.  The multitude of programs at UCLA enable students to join a program with similar goals in terms of their training which in turn allows the programs to be organized around these goals.

  2. Our Core Curriculum Provides Rigorous Computational Training. Our core courses are structured in the style of a quantitative Ph.D. program, complete with rigorous training requirements that are met through homework assignments, exams, and course projects. The philosophy behind our courses is to teach fundamental concepts in computation and use Bioinformatics to explore these concepts.

    For this reason, our core courses are rigorous enough to satisfy course requirements in quantitative Ph.D. programs at UCLA, including those for the Computer Science and Statistics graduate programs.  Bioinformatics core courses are taught and administered by faculty who have appointments in these quantitative departments.  Six of the courses are administered by the Computer Science Department, and one by the Statistics Department. Our rigorous core curriculum appeals to students in these programs as well as students in the Bioinformatics Ph.D. program. In fact, the majority of students enrolled our core courses are from quantitative graduate programs. This diversity of academic disciplines brings to these courses a high level of engagement and creativity.

  3. Substantial Didactic Training in Bioinformatics. Similar to a traditional quantitative sciences training program, our program offers a full load of Bioinformatics Courses. Our program includes five core courses that we strongly recommended students take during their first year.  These courses are: Introduction to Bioinformatics (Chris Lee), Algorithms in Bioinformatics (Eleazar Eskin), Methods in Computational Genomics (Jason Ernst and Bogdan Pasaniuc), Statistical Methods in Bioinformatics (Jessica Li), and Computational Genetics (Eleazar Eskin).

    In addition, students are encouraged to take during their second year Machine Learning in Bioinformatics (Sriram Sankararaman) as well as the multiple offerings of Current Topics in Bioinformatics (rotating faculty).  The Current Topics courses cover relevant issues such as Data Mining in Bioinformatics or Advanced Computational Genetics.  We designed the coursework for the UCLA Bioinformatics Ph.D. program so that students can take many skills-building courses comparable to those offered by a traditional quantitative science program.

  4. Rotation Program. Upon entering a Ph.D. program, students typically do not yet know whose lab they want to join. For this reason, we adopt a rotation program styled after typical Biomedical training programs.  Here, students undertake three 10-week rotations; one rotation during each of the three academic quarters of their first year.  Students use a rotation to try out a lab, and decide on a lab to join by the end of their first year in graduate school. Secondary, but important goals of the Rotation Program, are to develop diverse research skills, and to develop a collaborative network that may benefit the doctoral research project and career development.

  5. Seminar Program. An important aspect of Biomedical training programs is the informal training provided during seminars and journal clubs. The UCLA Bioinformatics Ph.D. program leverages informal training with a seminar that students are required to attend for the first two years of the program.  In fact, the weekly Bioinformatics Seminar series has become a key focal point of the UCLA Bioinformatics community.  Students also organize an annual overnight retreat where they share and get feedback on their research.

  6. Research Oriented Written Qualifier. Every Ph.D. program requires completion of a written qualifying exam, which typically occurs after coursework is completed. In traditional biomedical science programs, this exam is often preparation of a grant proposal in a topic of the student’s choice.  In traditional quantitative science Ph.D. programs, this requirement is often a challenging written exam covering topics in coursework. More recently, quantitative Ph.D. programs have abandoned the written qualifier and replaced it with an exam where the students write a paper demonstrating their research skills.

    In the UCLA Bioinformatics Ph.D. program, we have adopted such an exam.  After completion of first year courses, and faculty approval of their project proposal, students are given a one-month period to work independently on the project and to submit a written research paper reporting their results. Faculty in the program review the resulting papers. Although these projects are often small in scope because of the exam’s time constraints, the resulting papers are required to exhibit: 1) high quality in writing, 2) contextualizing the project within existing research, 3) supporting conclusions with chosen experiments, and 4) logical flow of the arguments in the paper.  The idea behind the exam is not to weed out students who cannot pass it, but to set an objective bar for achievement that the students can attain.

Just as Bioinformatics is an interdisciplinary field that combines methods, data, and theories from different academic traditions, the UCLA Bioinformatics Ph.D. is earned in an interdisciplinary program that combines aspects of the training cultures of quantitative and biological sciences. Our unit is a new kind of program that has been specifically designed to administer a rigorous, cross-sectional training in methodology development.

 

UCLA Bioinformatics: The Philosophy of the Undergraduate Program

Bioinformatics is an important interdisciplinary research area with tremendous opportunities in graduate training and industry employment.  Yet, few academic institutions offer undergraduate programs designed to prepare students for opportunities in Bioinformatics.

The UCLA Undergraduate Bioinformatics Minor is an academic program established in Fall 2012 at UCLA.  Undergraduates in any Major can obtain a Bioinformatics Minor by completing an additional 8 courses. Since Fall 2012, approximately 80 students have joined the Minor program. These students represent Majors in over a dozen UCLA departments, including: Computer Science; Chemistry; Molecular, Cell, & Developmental Biology; Microbiology, Immunology, and Molecular Genetics; Ecology and Evolutionary Biology; and Computational and Systems Biology.

Over 45 faculty specializing in computational and experimental biology are associated with the Bioinformatics Minor, spanning the fields of biology, mathematics, engineering, and medicine. Course offerings from more than 12 unique departments allow the Minor program to encompass the breadth of the growing Bioinformatics field.

Here we describe the principles and philosophy that guided the design of our Minor.

  1. Our Core Bioinformatics Courses Teach Interdisciplinary Computation. The foundation of our program is the cluster of three integrated core courses in Bioinformatics. These courses are truly interdisciplinary; they satisfy elective requirements in multiple departments and recruit students from different Majors to the Minor program. These core courses build upon the philosophy that students must first learn fundamental concepts in computation in order to later explore problems in Bioinformatics.  These courses offer basic skills and appeal to many students beyond those interested in Bioinformatics.
  1. Rigorous Background in Computation. To be successful in Bioinformatics, students must have a solid background in both computation and Biology. Our core courses require as prerequisites a substantial background in computation and statistics. To enter the Minor, we require that students have completed one year of programming and one upper division Statistics course.  To complete the Minor, our students take Linear Algebra and one upper division course on Algorithms taught by the Computer Science or Math Department.  Our students also take a Molecular Biology course taught by the Life Sciences Department. We believe that it is important for faculty in Computer Science and Program in Computing to teach programming, and for faculty in the Life Sciences to teach Biology. Further, it is important for students to take the same programming classes as do their peers in Engineering majors, and for students to take the same Biology classes alongside their peers in Life Sciences.
  1. The Bioinformatics Minor Builds upon the Students’ Major. Every student graduating from UCLA with a Bioinformatics Minor also completes an academic Major program. While we do adjust the Minor curriculum to help students efficiently complete both their Major and Minor requirements within 4 years, each of our graduates has exactly the same amount of training in their Major as fellow Majors who are not in the Minor.  This avoids a common pitfall in interdisciplinary education: students only receive a superficial background in each academic area.
  1. Bioinformatics is a Research Oriented Field. Our Minor is closely integrated with our undergraduate research program, which places students in the labs of Bioinformatics faculty. Most of the Bioinformatics Minors at UCLA are working in a research lab.  Undergraduates are strongly encouraged to engage in research. The Minor allows for a substantial amount of research credits, an allowance that helps students complete their Major and Minor requirements in four years.  In addition, many of our undergraduates participate in the Bruins-in-Genomics Summer (B.I.G. Summer) program or similar undergraduate education experience summer programs.
  1. Bioinformatics is an Increasingly Diverse Field. The core courses in Bioinformatics are designed to be interesting and accessible to students from a wide variety of educational backgrounds. Each course typically has enrollment approaching 100. Far more students who are not in the Bioinformatics Minor take these courses as electives to fulfill their Major requirements. Student enthusiasm is high for these accessible interdisciplinary courses that combine computational sciences and Biology. We find that this approach boosts broader undergraduate engagement in the field and encourages students from traditionally underrepresented groups to pursue research, graduate school, or careers in STEM fields.
  1. Let Excitement Foster Program Growth. Bioinformatics is an exciting area, and specialized training is critical for the next generation of biomedical researchers. However, undergraduate Bioinformatics programs, when offered by a college or university, are typically quite small.  Such programs are often limited in size and engagement as students are unaware of the field or become aware of Bioinformatics late in their college career. We strategized the Bioinformatics Minor program at UCLA specifically to attract students at any stage of their college career and to maximize curricular flexibility so students can easily complete Minor requirements. Many students are attracted to the Minor when they enroll in Bioinformatics core courses to fulfill elective requirements for their Major; some develop a keen interest in the field and then join the Minor. Even students who are unable to complete all Minor requirements benefit from our program; they complete key coursework and join a research lab, gaining knowledge and experience crucial for gaining employment or admission to graduate school.

Our current goal for the Bioinformatics Minor is to graduate 50 students per year.  We hope that 10 to 20 of them will enter graduate studies in Bioinformatics.  We are not there yet, but are growing. This year, around 10 graduates applied to Ph.D. programs in Bioinformatics.  Many of our students recently began or are applying to Ph.D. programs in Bioinformatics and related areas.  We expect that they will do very well in the admissions process and have great backgrounds for starting Ph.D. study in Bioinformatics.

bioinformatics-minor-graphical-element-the-minor

Read more about the Bioinformatics Minor on the official website:
http://bioinformatics.ucla.edu/undergradute-bioinformatics-minor/

Check out a list of research opportunities available for undergrads at UCLA:
http://bioinformatics.ucla.edu/undergraduate-research/

Learn more about 2016 undergraduate research and B.I.G. Summer activities at ZarLab:
zarlab.cs.ucla.edu/b-i-g-summer-in-zarlab/

Applications to the 2017 B.I.G. Summer program are due January 27:
http://qcb.ucla.edu/big-summer/