Bioinformatics training through metagenomic sequence annotation.
We have taken advantage of the increasing amount of metagenomic data publicly available to develop a teaching environment in which undergraduate students are given the opportunity to “turn data into knowledge”. Called Annotathon, this internet teaching platform fosters team work and guides apprentice annotators through each step of in silico analyses, from ORF identification to functional and phylogenetic classification. Generating raw results is an integral part of the exercise, but heavy emphasis is put on their interpretation and critical assessment.
The online format is ideally suited for student involvement outside class, whilst allowing instructors to provide annotators with continuous feedback. An iterative evaluation cycle allows students to respond to constructive criticism and produce enhanced versions of their annotations. A student exit survey shows that, although challenging, the study of as yet non-annotated sequences is a major incentive for 75% of participants. The success of this approach relies on the following:
i) Learning by doing: bioinformatics is best introduced by first hand experience; theoretical considerations are easier to grasp once truly familiar with the tools.
ii) Learning by repetition: successively analysing several sequences gives the students the opportunity to experience a wide range of situations, e.g. “twilight zone” similarities, or “garbage in, garbage out” phylogenetic tree reconstructions.
iii) Learning from constructive criticism: giving students the opportunity to correct themselves results in noticeable progression over time.
502 students have taken part in the Annotathon over the past three years, analysing a total 2.3 Mb of ocean microbial DNA, representing 9500 hours of cumulative annotation. Interestingly, student hand crafted annotations are in very close agreement with those obtained by automatic annotation pipelines. Based on these observations, we call on our colleagues to consider joining an initiative to create a worldwide “grid of undergraduate annotators”. Tentatively, as more teams join in, this could lead to an educative annotation community with room for scientific contribution. The open source online Annotathon environment (http://annotathon.univ-mrs.fr) is offered as a basis for this collaborative initiative.