National Survey 2018

This year, for the first time, COMBINE ran a national survey of students. The goal of the survey was to learn more about the Australian bioinformatics and computational biology community, specifically:

  • Who is studying bioinformatics and computational biology in Australia
  • Their academic backgrounds and skills
  • The areas of research and technologies they are working with
  • The support they receive from their supervisors, institutes and universities
  • What issues they encounter during their studies
  • What COMBINE can do to support the student community

The survey opened in April and closed in June and we received over 80 responses. Thank you to everyone who completed the survey and congratulations to those that won a gift voucher in our prize draw.

Included here are some of the interesting results from the survey. When interpreting what is shown here it is important to remember that this is a fairly small and selective sample. As a result the data is more a reflection of the COMBINE community than the wider student community, however, we still believe this has been a useful and informative exercise. We also believe that this is a valuable opportunity for students to see what others’ experiences are like and contrast them with their own.

Who completed the survey?

We received 83 responses to the survey, after removing a couple that were duplicated or incomplete. The majority of these were from full-time PhD students, however there were some other course types and a few part-time responses. The sample was also biased towards female and domestic students.

Perhaps the most obvious and unexpected demographic is in the distribution of student locations, particularly the high numbers from Victoria and the low numbers from NSW. This is likely to be at least partly a result of COMBINE’s history as we were based in Melbourne for a long time and have had many Victorian members of the executive. We are hoping to address this in the second half of the year by running more events in Sydney. If you are based in NSW and would like to get involved we would love to hear from you, particularly as the COMBINE Symposium will be held in Sydney in 2019.

We also asked students about their academic backgrounds. The majority were coming into bioinformatics from biology, but male students were more likely to come from a computer science or software engineering background.

Other backgrounds included pharmacy, biotechnology, psychology and various types of engineering. It is unsurprising to see many backgrounds as bioinformatics has traditionally drawn from many fields, but it will be interesting to see if this changes as more computational subjects are included in undergraduate degrees.

Bioinformatics and computational biology can require a diverse skill set including aspects of biology, computer science and statistics. We asked students to rate how competent they felt at a variety of skills on a scale from 0 (lowest) to 5 (highest).

Overall we saw quite a spread of ratings for all skills. Female students were more confident in their genetics skills while male students were more comfortable with programming and software development. International students rated themselves more highly at many of the technical skills including computational algorithms, linear algebra, programming and software development.

Money can be a source of stress for many students and we were interested in how bioinformatics students supported themselves. In general scholarships (university and government) and part-time work were the biggest sources of financial support, but unsurprisingly this is different when factors such as course type or residency are considered.

Domestic students are much more likely to be supported by government scholarships or part-time work while international students typically rely on university funding. Master’s students rely more on part-time work and parental support, but worryingly half also responded that they supported themselves using their savings. We neglected to included government welfare as an option in the survey but this was recorded in other responses, along with scholarships from other funding bodies.

What are students studying?

Of the 83 responses, 20 were currently completing coursework classes. The majority of these were Master by coursework students but they covered all levels from Bachelor to PhD.

Coursework students were enrolled in one to four classes in the first semester of 2018 with between one and 30 hours of classes a week. There was considerable variation in how classes were assessed with students having up to 16 assignments in a semester and up to six exams.

We asked students what percentage of their current classes contained biological, statistical or computational content. One class may fit into two categories, for example a statistics subject might teach statistical theory but also how to implement it computationally, meaning that the total could come to more than 100 percent.

We found that classes are fairly evenly spread through the three disciplines, although there are less interdisciplinary subjects than we expected. This may represent a truth about the fragmented nature of universities or it could be related to how these questions were worded.

Detailed comments provided by students give more insight into the coursework experience, particularly some of the negative aspects. One student commented that their Master’s course has “No support. No prerequisites” and was “Badly designed”, while others focused on assessment: “Too many assignments. Interrupts meaningful learning”. Other students felt they were not being adequately prepared for the field: “… the coursework is geared towards experimental projects with no bioinformatics or statistics unfortunately”. One detailed response touches on many of the difficulties for coursework students, particularly those from biology backgrounds:

“The pre-requisite mathematics knowledge for the biology stream are far lower than what is expected from all of the probability & statistics courses. There are no bridging courses, and no support for students who are having difficulty in these courses. Most biology stream students who have taken these courses have struggled with the content, even when they put 90% of their time into studying for these courses at the expense of others. Given the grades necessary to enter a PhD program with a scholarship, it feels like being set up to fail.”

This comment raises some considerable concerns and suggests that universities continue to grapple with preparing students for modern, data-focused biology. We expect that universities will continue to develop and improve
interdisciplinary courses and integrate more statistical and computational skills into undergraduate biology courses, hopefully leading to less of these experiences in the future.

What are students researching?

The majority of responses were from students who are currently completing a research project (74 of 83). These projects cover a range of research focuses. As we would expect of bioinformatics students the majority are involved in data analysis, with methods development and data visualisation also common. Of the research topics, gene expression and genomic variants were most common but there was quite a range of other focuses in student research.

Short-read sequencing technologies including RNA-seq and whole genome sequencing were the most common sources of data. Other technologies not given as options include ChIP-seq and ATAC-seq.

A core set of four programming languages were used by students in their research (R, Shell, Python and Git), with 80 percent of students using R. Other tools are used by some students but these four are clearly the current dominant programming tools.

While many students use the same sets of computational tools there is considerable variation in the percentage of their projects that are computational.

Male students and international students were more likely to have purely computational projects while female and domestic students showed a greater spread. This may be related to students’ academic background with female and domestic students more likely to have a biology background and the related lab skills.

We also saw differences between students as to who they were collaborating with for their research.

About twice as many male students were collaborating with other groups at their university or institute while more female students collaborated with other students.

Who are their supervisors?

Supervisors can be just as crucial to the success of a research project as the students themselves. The majority of research students had either two or three supervisors with primary supervisors being responsible for two-thirds of supervision on average.

The majority of primary supervisors (77 percent) were male. Female students were slightly more likely to have female primary supervisors but were also more likely to have male secondary supervisors. It is likely that the prevalence of male supervisors is a result of the gender imbalance at higher levels of academia rather than any preference from students.

We asked students to rate their supervisors on a range of qualities that can lead to good student-supervisor relationships and successful projects.

Female primary supervisors were rated more highly on most qualities, with the exception of subject matter expertise. It is important to note that these ratings are subjective and may reflect the biases of students. Overall these ratings are high, suggesting that most supervisors are generally doing a good job. However some students had issues, particularly with supervisors without computational backgrounds. For example, one student mentioned their supervisor had a “…lack of expertise in the field combined with lack of collaboration with other groups/researchers, and even not allowing me to consult other experts in the field …”, and another “Supervisor has no computational background so getting guidance on this area is challenging”.

Other students had mixed experiences with their supervisors, highlighting why it can be important to have more than one:

“My primary supervisor is truely wonderful. She’s super supportive, engaged, and even though she is busy, always makes time for me. Although I have a secondary supervisor, I have had maybe 2 meetings with her my whole PhD (2.5 years in) and she hasn’t attended any of my official events (confirmation, review, presentations etc.).”

What support do students receive?

We asked coursework students how supported they felt by their university and student cohort.

Significantly more female students answered yes to both questions. While all female students felt that they were part of a student cohort this was true for less than 75 percent of male students.

How supported research students feel can depend more on their supervisor and research group than their university.

Most research students felt like they were part of their research groups, one student commenting “I have a fantastic relationship with my supervisor and fellow lab-members. I couldn’t be happier with my research environment”. However, for the minority of students that felt differently, this seemed to be related to the cohesiveness of their groups with students explaining that their research group was “… not tight or supportive enough. Everyone working independently with basically no interaction …” and “It’s very exclusive, and there is close to zero research interaction in the group (everyone works individually)”. These feeling could also be related to the 15 percent of students that are alone in their research groups.

We also asked students more broadly about their sources of support outside their supervisors. Overall 73 percent of students felt supported by their research group, 65 percent by their institute and around 55 percent by other people or students in their lives. These values are lower than we would hope to see and ideally every student would benefit from all these sources of support. Most worryingly 8 percent of students said they had no source of support outside their supervisor.

Female students generally felt more supported, as did domestic students and those with female primary supervisors. Students associated with, or based at, a research institute were more likely to feel supported by their research group than those based purely at a university.

One way that support can come from supervisors is through meetings with students. We asked how often student have formal (scheduled meetings with the research project as the main topic) and informal (getting coffee, asking questions, coming by a desk etc.) interaction with their supervisors.

Most male students had formal meetings once a fortnight and informal interactions a few times a week. Female students were more variable, meeting both more and less frequently. Meeting frequency may depend on workload and preference as this comment notes: “If I have a lot of trouble and questions I might see them everyday but then there are periods where I maybe see them once a month. I really like the freedom of that”.

Access to appropriate facilities is also vital for students to successfully complete a research project. We are pleased to see that the majority of students are provided with desks and computers as well as access to servers or high-performance computing that can be essential for many bioinformatics projects.

Funding to attend conferences can also be important, particularly for PhD students hoping to continue in academia and thereofore needing opportunities to present their work and meet people in the field. Most students had access to travel grants with an application through their institute or university, but those based at research institutes had greater access to top-up scholarships.

What can COMBINE do to help?

The last section of the survey was designed to find out how COMBINE can help to improve the student experience. We started by asking students what previous experience they had had with COMBINE or ABACBS.

Students were most likely to connect with COMBINE and ABACBS via mailing lists and social media and around a quarter had attended the COMBINE symposium or ABACBS conference.

About half of students reported they were paid ABACBS members and we asked what they thought the benefits of membership were.

Discounts for events and conferences received the most positive responses as demonstrated by these comments: “I think the most important thing is decrease the registration fee of conferences” and “… provide more funding support to attend symposium”. Access to prizes, awards and travel bursaries was perceived to be more beneficial by non-members than members. International students were generally less positive about membership benefits, with the exception of it looking good on a CV.

When asked what future COMBINE events students would like to see, workshops were the most popular option, along with the COMBINE Symposium. The general programming and introductory bioinformatics workshops we have run in the past have been successful, but clearly there is demand for more as evidenced by this comment: “I would love to see more beginner workshops, as I really benefitted from a workshop ran by COMBINE when I first started sequence analysis”. Statistical workshops are not something we have done in the past but are something to consider if we can find a format and content that works.

Local social nights and student seminar series are the most common types of events that COMBINE runs but these were of interest for less than 35 percent of students that completed the survey. We believe these events are important for building a student community but there is more we can do to make them more attractive for more people. For example, one comment suggested that COMBINE should “… hold monthly bioinformatics meet up in each city and the program includes allocating desks/tags/etc for different topics (eg. DNA seq, RNA seq, etc.) for asking questions, networking, etc.”.

Other suggestions included mentoring programs, advocacy for student talks and travel funding and a PhD award (stay tuned!). One comment asked us to provide “… a framework where students can connect with others for help and support with aspects of their project, e.g. questions on programming (language specific groups), troubleshooting advice, beta testers, etc.”. Earlier this year we opened up our Slack workspace for any students to join and we hope it can become exactly that kind of platform, as well as giving students more opportunity to get involved with COMBINE.

On a final note, we really appreciated receiving comments like this, which we think embodies exactly what COMBINE is trying to achieve:

“COMBINE is the greatest. I value the student-driven focus and identity of COMBINE, as well as its role as an ABACBS sub-committee which enables cohesion with the broader community whilst allowing students a space to develop community leadership skills. Keep doing what you are doing!”