Bioessays Online Submission Training

Please, wait while we are validating your browser

Abstract

The mountains of data thrusting from the new landscape of modern high-throughput biology are irrevocably changing biomedical research and creating a near-insatiable demand for training in data management and manipulation and data mining and analysis. Among life scientists, from clinicians to environmental researchers, a common theme is the need not just to use, and gain familiarity with, bioinformatics tools and resources but also to understand their underlying fundamental theoretical and practical concepts. Providing bioinformatics training to empower life scientists to handle and analyse their data efficiently, and progress their research, is a challenge across the globe. Delivering good training goes beyond traditional lectures and resource-centric demos, using interactivity, problem-solving exercises and cooperative learning to substantially enhance training quality and learning outcomes. In this context, this article discusses various pragmatic criteria for identifying training needs and learning objectives, for selecting suitable trainees and trainers, for developing and maintaining training skills and evaluating training quality. Adherence to these criteria may help not only to guide course organizers and trainers on the path towards bioinformatics training excellence but, importantly, also to improve the training experience for life scientists.

bioinformatics, training, bioinformatics courses, training life scientists, train the trainers

INTRODUCTION

The hunger for bioinformatics training courses arose around the mid-‘80s, following the appearance of the first databases and software tools for analysing protein sequences and structures. Back then, there were no easy ways to disseminate such resources, and computer facilities in most life science laboratories were crude or non-existent. Novel data-distribution mechanisms had to be invented to ensure that the new resources were reaching their target audiences. The European Molecular Biology Network (EMBnet), for example, pioneered the distribution of the EMBL Data Library [1] from the EMBL in Heidelberg to national data centres holding government mandates to provide access to this and other bioinformatics databases and tools to their local communities [this model is now being adopted, on a much larger scale, by ELIXIR (http://www.elixir-europe.org/), a pan-European endeavour to provide a sustainable infrastructure for biological information, and which will generate even greater training needs] [2]. Such distribution networks solved many problems for data providers but demanded a certain level of end-user computational competence: first, to be able to login to a remote, centralized site; second, to be able to find and access the relevant databases or software tools on the remote system; and finally, to be able to export any results back to the local computer. The web did not exist, and most of these skills were the preserve of just a few self-taught “informaticians”, who were comfortable with arcane Internet communication protocols and search engines, such as Gopher [3], WAIS [4], Archie [5], HASSLE [6] and so on. Training courses became essential to allow life scientists to overcome the technical hurdles, their focus necessarily being on “how to access” bioinformatics tools and resources.

In the early ‘90s, the advent of intuitive graphical web browsers shifted the goal posts. For the first time, databases and software could be accessed instantaneously via customized web interfaces. These were designed to be as easy to use as possible, often “hiding” some of the more technical details and parameters behind “advanced” options that most users never dared to explore. However, as web technologies moved on, the database and software interfaces accreted greater degrees of functionality and, ironically, became harder and harder to use. Consequently, a new breed of training courses on “how to use” bioinformatics tools and resources was born.

The past decade has witnessed another shift: the industrialization of laboratory techniques has revolutionized the pace of data acquisition, computers are now standard laboratory equipment, and both the computational competence and computational requirements of life scientists have increased accordingly. The scale of data generation, today, is daunting—laboratory automation has made it possible to gather data first and to formulate hypotheses later. Indeed, such ‘data-driven science’ [3, 7] is now commonplace. Thus, more than ever before, researchers want to know, ‘How should I analyse my data?’, ‘How do I get the best out of this or that computational tool or resource?’, ‘What do my data mean?’ or even, ‘What is my hypothesis?’. Bioinformatics training courses are having to adapt to meet these new needs, but the pace of change has been swift, creating new challenges for course organizers and trainers, and ultimately also for trainees—how, for example, can they be certain of receiving the best, most excellent training?

WHAT IS TRAINING EXCELLENCE?

The focus of training is the trainee. They ultimately judge training excellence not just in terms of how they perceive a particular training event, but also in terms of the impact this has on the development of their skills in the long-term. Excellence in training could be generally defined as the ability to deliver appropriate training in response to a particular demand, providing high-quality, up-to-date content and satisfying the expectations of trainees, of trainers and of the organization providing the training.

For several years, the Bioinformatics Training Network (BTN) [8] has provided a forum for bioinformatics trainers to share their experiences, to identify common challenges [9] and to agree on common working practices [10]. From these shared experiences and from round-table discussions on what can be understood by training excellence and how it might be achieved, four repeating themes have emerged that are generally applicable to the delivery of successful training for life scientists and beyond: (i) understanding the needs of trainees; (ii) ensuring that the training provided is suitable for a given audience; (iii) ensuring that a quality-assurance process is in place; and (iv) defining a sound organizational framework. These four aspects encompass many related facets; excelling in all is the key. From our collective perspectives and experiences, we prepared and made freely available, as a deliverable of the EU SLING project, an extensive document entitled, ‘Bioinformatics training for life scientists: guidelines for best practice’, based on what we believe ignites excellence in training: iterative performance of training events, assimilation of what did and did not work and feeding this information back in a dynamic feedback loop. Here, we present a summary of our discussions and invite all those in life science research and education to contribute to our on-going dialogue on how best to create a robust and sustainable foundation for bioinformatics learning, education and training.

Identifying training needs

A training need arises when an individual is unable to perform a task adequately, or cannot perform it to a sufficiently high standard. Currently, significant training needs in the life sciences have arisen from the rapid advances in high-throughput data-production technologies, coupled with the volume and complexity of the data these are producing; the pace of change is so great that there is a growing lack of exposure to the tools and technologies for handling, retrieving, analysing and interpreting these data, and a dearth of understanding in how these might contribute to biological discovery. Courses addressing such needs are more likely to succeed if their target audiences are sufficiently specific to be able to narrow the focus to aspects that are relevant to the participants’ own research projects, to their level of background knowledge and to their technical experience with bioinformatics tools/resources. For example, from the technical standpoint, an important (but often over-looked) consideration is trainees’ familiarity with the Unix/Linux command line, R, etc—especially in courses that cover next-generation sequencing (NGS) data analysis. It is crucial to recognize the need for experience with Unix/Linux, either as a course pre-requisite or as a training need that can be addressed at the start of a course. Gathering such information from candidate participants in advance helps to identify this kind of training need [10].

Set learning objectives

Training needs should be perceived as such from both sides—by trainers as well as trainees. Therefore, explicitly mentioning the learning objectives (LO) of a course, or of a specific section of it, is strongly recommended. An LO is a clear statement of what the trainee(s) will be able to do as a result of the training, to what standards and under what conditions. LOs should be mentioned in the course description and designed in tune with participants’ backgrounds and capabilities. LOs should always be formulated in terms of competencies, using verbs like ‘reproduce’, ‘apply’, ‘predict’, ‘compare’, rather than ‘know’. This is because the former abilities can be translated directly into practical tasks and exercises, which represent essential tools to achieve LOs, whereas knowledge is related to principles, and it is usually acquired more indirectly through long-lasting experience or university courses.

Matching training provided to audience

Selecting suitable trainees

Most training programmes and individual events are planned with the assumption of a particular training need in an, as yet, unknown audience. Prospective applicants will need to apply under one of a variety of possible mechanisms, from first-come-first-served to specific selection procedures. Matching the suitability of trainees to the training offered becomes a significant challenge in itself. For example, two potential trainees may need to know about NGS-data analysis: their end goals may be the same, but if one is a biochemistry researcher with an MSc in computational biology and the other is a clinical geneticist, they are likely to need to take different routes to achieving them. Therefore, whenever possible, it is recommended to define selection criteria that allow collation of applicant information, regarding: (i) relevance of the course topic to their scientific needs; (ii) their expectations about the course (e.g. are these realistic?); (iii) the suitability of the scope of the course to their career stage (e.g. are they well matched?); (iv) their fulfilment of course pre-requisites (e.g. can they program in Perl?). This information can be obtained by including a brief questionnaire in the course application form. When it is not possible to collect previous information about applicants, it may anyway be useful to do it at the start of the course, to have the possibility of adapting the teaching accordingly.

Of course, despite having followed these recommendations, it is still possible that a selected group of trainees may not fit a course perfectly, or may not be satisfied by it. This can happen for various reasons: not all trainees are capable of learning everything—some aspects of a course may simply be too difficult for them; some trainees might have been obliged to apply to a course to plug a perceived skills gap but find the course pitched at the wrong level; others may have been pressured to apply to fulfil the needs of their project, but find they have no genuine interest in many (or all) of the course topics. Situations like this rely on trainers’ sensitivity to detect these circumstances and to pay special attention to motivate such participants, e.g. by involving them in the solution of exercises before a class or giving them specific, tailored assignments, such as wrapping up at the end of the day or leading a brainstorming session.

Identifying appropriate trainers

Good trainers not only have appropriate subject knowledge but also good pedagogical and andragogical skills, are conscious of individual learning styles and paces and have the ability to ensure that participants interact and maintain their interest. Once the need for a specific training course has been identified, the organizer has to decide who will teach it. Unless the host organization has qualified trainers available, this is not an easy task. Indeed, there are no resources providing lists of recognized or accredited bioinformatics trainers, and most recruitment still occurs through personal knowledge of specific individuals, regardless of whether better trainers exist. A good candidate trainer is someone who is both expert in a topic and has experience of teaching it, whether in academia or in bespoke training courses. Generally, as the approach to short courses is fundamentally different from academic teaching, trainers with specific short-course experience may be more suitable than university professors. However, many good trainers have experience in both short training- and longer educational courses, and their teaching practice may be the richer for it.

Course organizers (individuals or institutions) represent a possible source of information when seeking appropriate trainers. Furthermore, for a course on a specific bioinformatics resource (database or tool), advice may be sought from the resource developers: often, they are able to provide specialized trainers or to organize courses themselves.

In an effort to make trainer selection easier, organizations like GOBLET (Global Organisation for Bioinformatics Learning, Education and Training) are working to collect and make available the names and competencies of experienced bioinformatics trainers without making value judgements. How to develop databases of, and effective rating systems for, trainers is currently a hot topic.

Preparing the training

Bioinformatics training should be flexible to accommodate different types of content, course duration and trainee-learning speeds and skill levels. A common theme is the need to select a digestible amount of content and to prepare bite-sized chunks of training. Choosing appropriate teaching methods and preparing course materials are also part of the training groundwork.

Choosing the course format

Choosing the right format depends critically on striking the right balance between course duration, level and participant backgrounds. In deciding on a training format, it is worth considering: the trainer-to-trainee ratio, the number of participants, the time available, the facilities available and the experience and expectations of the trainees. Table 1 summarizes five formats commonly used by the authors and their pros and cons.

Table 1:

Pros and cons of different training formats

Course format Pros Cons 
Lecture + PC practicals Easy to structure and prepare, even if done by separate persons. Lectures allow for easier face-to-face communication. A strong topic connecting both parts is necessary, otherwise the lecture may be perceived too theoretical. 
100% PC practicals Best suited to self-learning groups with lots of material, and if the trainer takes the role of a coach rather than an instructor. There is little room to cover extensive theoretical content. Direct communication may be limited because the PCs draw attention away from the other course participants. 
Seminar with PCs ready Groups of up to 10 people can switch between PCs and face-to-face teaching smoothly. Works best with a PC-free zone in the same room. Conference table with laptops also works. Difficult with larger groups. The PCs pose a distraction to some extent. 
Remote e-learning session No travel costs; potential to train large numbers of people. Requires highly motivated and independent trainees and well-prepared material. The plan is difficult to change on the fly. 
Blended learning (combined teaching approach) Potentially allows the disadvantages of all other approaches to be overcome. Higher investment for course organizers, requiring more planning. 
Course format Pros Cons 
Lecture + PC practicals Easy to structure and prepare, even if done by separate persons. Lectures allow for easier face-to-face communication. A strong topic connecting both parts is necessary, otherwise the lecture may be perceived too theoretical. 
100% PC practicals Best suited to self-learning groups with lots of material, and if the trainer takes the role of a coach rather than an instructor. There is little room to cover extensive theoretical content. Direct communication may be limited because the PCs draw attention away from the other course participants. 
Seminar with PCs ready Groups of up to 10 people can switch between PCs and face-to-face teaching smoothly. Works best with a PC-free zone in the same room. Conference table with laptops also works. Difficult with larger groups. The PCs pose a distraction to some extent. 
Remote e-learning session No travel costs; potential to train large numbers of people. Requires highly motivated and independent trainees and well-prepared material. The plan is difficult to change on the fly. 
Blended learning (combined teaching approach) Potentially allows the disadvantages of all other approaches to be overcome. Higher investment for course organizers, requiring more planning. 

View Large

Diversity of training methods

In face-to-face training, a plethora of methods can be used to deliver a successful course. In our experience, three golden rules apply: (i) the trainer should present content in an engaging way; (ii) the trainees should be stimulated to think actively during exercises; and (iii) interaction and discussion should be encouraged.

How can these rules be translated into specific actions and choices? Trainers are constantly in search of effective training methods; most want to find an optimal balance among the many available options: showing slides, promoting discussions and interactivity, solving exercises together, stimulating individual work, asking trainees to present a topic (‘flip classes’), organizing games, telling engaging stories, working in groups and so forth. None of these activities alone is a guarantee of success, and various stakeholders have suggested that the most effective balance is achieved through multimodal learning [11]. The question is, are some approaches more effective than others?

Many bioinformatics trainers use slide-based lectures and live demonstrations for information transfer and practical exercises to reinforce learning. This format is not necessarily optimal for teaching a new competency, as opposed to simply transferring knowledge. Examples of teaching methods used successfully in bioinformatics training courses include: (i) use of case studies (reduces the complexity of the subject and tells a ‘research story’ to which trainees can relate); (ii) provision of teaching materials, such as manuals, glossaries, tasks and questions (help trainees to learn independently and lift some of the burden from trainers); (iii) explanation of algorithms using simplified models supported by board games, role-plays or pen-and-paper implementations [12, 13]; and (iv) discussions in groups, and with the entire class (brainstorming, gathering pros and cons, panel discussions and so forth). Incorporating a variety of such methods helps address differences in trainees’ preferred learning styles and learning paces and is more likely to be effective than traditional approaches. The balance between different modalities, however, may depend on a trainer’s attitude and capability (for example, an engaging speaker may be more successful spending time presenting content than in making trainees work in groups).

Overall, learning is a complex phenomenon [14]. Experience suggests that the most effective training approaches combine several styles, which may vary from one trainer to another and from one audience to another, and should be adapted to the training circumstances. Attention, motivation and basic skill levels of trainees also play fundamental roles [10, 11, 13].

Creating a training plan

A training plan is a scheme of the content, teaching method(s), goals and time allocated for each phase of a training session (Table 2

0 thoughts on “Bioessays Online Submission Training

Leave a Reply

Your email address will not be published. Required fields are marked *