Transcriptome-profiling is the primary means by which researchers characterize what is happening at the molecular level. When a transcriptome profile is generated for a sample, it tells us which genes are active (being expressed) and at what amount. Comparing profiles among samples allows us to identify genes, and importantly the biological processes that are varying systematically, whether by disease, treatment or any other variable of interest.
As high-throughput sequencing has become more widely available and more cost effective, transcriptome profiling analysis is becoming a standardized tool for molecular characterization of differing biological systems.
Researchers who work with common model organisms such as mouse, zebrafish or nematode enjoy easy access to large swaths of tools and resources for analyzing and interpreting their experimental data. Specifically, because these organisms have been so deeply studied, we already know many (or even most) of the RNA transcripts that can be made. The availability of these known transcripts greatly simplifies the process of interpreting a new RNA-seq data set. In contrast, researchers who study uncommon (non-model) organisms are often required to build specific tools and consolidate or create their own resources. Research groups that work in isolation and with modest computational resources, may find successfully completing these tasks even more difficult.
Wanting to help overcome these obstacles, discussions with researchers in the Maine INBRE network (IDeA Network of Biomedical Research Excellence,) led us to launch the 2019 Maine INBRE Nonmodal Organism Transcriptome Analysis workshop, MINOTA for short. MINOTA 2019 was a one day, on-site affair, with students from across the state gathered on our Salisbury Cove campus, where they gained hands-on experience with the most commonly used tools and methods for building and analyzing transcriptomes.
Additionally, there was a focus on rapidly bringing participants up to speed on the basics of working on the command line (a vital skill in any fledgling bioinformatician’s toolbox). And, leveraging the power of Amazon Web Services, every student gained access to the robust computing capacity required to test their handiwork.
While the primary goal of the workshop was to explore the process of transcriptome analysis, we also aimed to cultivate a growing community that was linked by the necessity for these tools, and to provide a central resource for further education and developmental needs.
Students were also introduced to the importance of rigor and reproducibility in bioinformatics processes. These “best practices” are a set of standards that we (the Bioinformatics Core) adhere to so that our code, directory structure, and experimental output don’t resemble a plate of spaghetti (with no disrespect to the delicious dish!) It also ensures that we can reliably produce the same results, no matter when, or even where, we conduct our analyses.
Due to the Covid-19 pandemic, this year’s MINOTA workshop was markedly different, as the pandemic meant that an in-person event wasn’t possible. We took this radically different scenario as a challenge and creatively turned MINOTA into a five-day virtual event. The resulting multiple-day workshop was worked out well; transcriptome assembly (keep an eye out for a future blog post that describes this process in more detail) is computationally intensive, and even with high-powered computers it can take many hours or even days to complete. This year we were able to move from using “toy” problems (ones that could be completed in the limited time available) to having participants work through larger, more realistic data sets. More time also meant better learning outcomes, as students had the space to absorb what the materials and come back with any follow-up questions.
The course content was adapted to be designed for remote learning. Pre-recorded lecture materials so that the participants could view them in their own speed. Materials covered everything from the basics of command line file editing, to transcriptome assessment and annotation. Help materials and exercises to work through were provided online through web pages and short videos, and finally, twice-daily live video discussion sessions were held. Finally, all participants were connected through the MDIBL-developed and hosted LabCentral platform, which enabled ongoing discussions and help where needed for specific troubleshooting and specialized support.
The workshop was attended by 17 participants, including 7 teaching or research faculty, a research postdoctoral fellow, a core manager, 7 graduate students, and an undergraduate student. All five states in the Northeast IDeA region were represented, and participants joined the course from 12 different home institutions.
While this year’s challenges were new and unexpected, in the end, we believe that this forced us to grow and improve the MINOTA workshop in a way that will make future versions stronger yet. We look forward to hosting MINOTA 2021.