Attention Conference Presenters - please review the Speaker Information Page available here.
If you need assistance please contact submissions@iscb.org and provide your poster title or submission ID.
Short Abstract: TeSS (life sciences training portal, https://tess.elixir-uk.org) aggregates links to disparate training events and materials, scattered across ELIXIR Nodes and other training portals, and makes them centrally discoverable.
Aggregation is effected automatically, via custom-built scripts that scrape training metadata descriptors (title, description, prerequisites, goals, categories, keywords, contributors, audience, etc.) from participating websites. This task is made difficult by the lack of consistency in training pages’ structures and lack of agreement on training metadata descriptors.
In the past year, we helped initiate the BioSchemas Working Group, which aims to reach consensus among life-science communities in defining metadata standards for describing training resources. BioSchemas are extending specifications from Schema.org, reusing existing metadata standards. This is in line with Google and other major search engines, which already utilise Schema.org’s schemas for search optimisation.
We have also started to actively promote the use of ‘bioschemas’ among content providers. A first follower was GOBLET, whose training portal has embedded the new BioSchemas-proposed mark-up within their training pages. Ultimately, our goal is for all content providers to adhere to the agreed schemas, making the job of search engines more efficient, and scrapers redundant.
Our next immediate task is to increase the content from ELIXIR Nodes, and to further promote take-up of the evolving ‘bioschemas’. At the same time, to make further efficiency gains and broaden the scope of the resource, we are partnering with the emerging US BD2K’s Training Coordination Centre, which has a similar remit to that of TeSS, sharing expertise, training resources and technical know-how.
Short Abstract: The Moore Method was originally developed by R.L. Moore to teach advanced mathematics in the college setting. There have been many adaptations of the Moore Method, under the broad term Modified Moore Method (M3), which are now classified as a variant of inquiry based learning (IBL). Despite the growing popularity of M3, it is rarely applied beyond mathematics. At Iowa State University, we designed and taught an “Introduction to Bioinformatics” undergraduate course using M3 for the first time during Fall semester 2015. The class size was small (n=12), and students all had a background in the natural sciences, most in the biological sciences. Students had little to no formal training in computational sciences. During the 16-week course, students learned to: 1) work on a remote Linux server, 2) read and write Python code, 3) tackle classic bioinformatics problems, and 4) solve current bioinformatics problems with available tools. As with all M3 courses, learning objectives were met through carefully designed questions given to students prior to each class session. Class sessions were completely led by students (i.e., reversed classroom) presenting solution to the assigned questions. The application of M3 to our course has led to several desirable student outcomes: 1) engagement and ownership of the course material, 2) development of a strong sense of community, and 3) uniform learning outcomes. One of the difficulties we experienced with applying M3 was the creation of the course material. It was tough to create questions that were challenging enough without overwhelming the students.
Short Abstract: The beginning of the 21st century has seen an increased data deluge implying that high-performance computing is becoming essential to process the information in a timely fashion. For example, executing large number of independent tasks or tasks that perform minimal inter-task communication in parallel is a common requirement in many domains. Despite the data globalization, the data / algorithmic complexity, heterogeneous data sources and nomenclatures, impractically excessive runtimes, etc. make the data exploration and mining a huge challenge. It is obvious that comprehensive studies to understanding complex biological organisms will require computational processing power and alternative paradigms for efficient data integration, replication and organization.
At UCR’s high-performance computing|bioinformatics facility, we have developed a series of parallel-, multi-core-CPU-based open-source pipelines for large-scale -omics data analysis, which enables efficient and parallel analysis of multiple datasets in a short time. Our unit has been contributing to different research projects by providing, (i) Support and expertise in programming and advanced data analysis, focusing primarily on high-throughput genomics technologies including microarrays, genotyping and next-generation sequencing (RNA-seq, ChIP-seq, SNP-seq), (ii) Virtual server environments, secure and public web portals, and a large suite of open source applications, (iii) Instruction of hands-on workshops on a wide informatics topics, (iv) Custom data analysis and consultation services, and (v) Establishment of research collaborations with experimental scientists from different departments. The facility is open to all researchers, UCR as well as to external users anywhere in the world. Detailed information about the services, research, manuals developed, etc. is available at http://bioinfo.ucr.edu/.
Short Abstract: The Global Organisation for Bioinformatics Learning, Education and Training (GOBLET: http://mygoblet.org) was established to provide a global, sustainable support structure to foster international communities of bioinformatics trainers and trainees. The activities of GOBLET are carried out through committees, which have independent but overlapping focus areas. The Learning, Education and Training (LET) Committee primarily focuses on providing resources for bioinformatics trainers. Here we describe some of the recent activities and resources developed by the LET Committee: (i) A set of consensus descriptors for training materials to ensure that materials are consistently described with a minimum, standard amount of information. This brings a strong improvement in discoverability, shareability and traceability of training materials. (ii) The development of core competencies together with the ISCB Education Committee, and how these can be used to elaborate bioinformatics curricula and training materials appropriate for different audiences. (iii) Our e-learning activities in bioinformatics from the perspective of discoverability of existing e-learning materials and the development of new materials. For these activities we partnered up with other networks and organisations with similar goals.
Short Abstract: Measuring training impact is an important yet challenging element of providing a coherent and quality training programme. There is much discussion around the “best” methods for measurement, but consensus is yet to be achieved – a one size fits all approach may not be appropriate.
At EMBL-EBI we run a wide-ranging training programme covering data analysis topics across the life-science spectrum. Until recently our focus has been on collecting feedback immediately post-course, providing insight into short-term impact. In January 2016 we embarked on a programme of longer-term assessment, approaching trainees 6 months post-course to try to capture impact measurements.
Definitions for impact will inevitably vary between programmes, but for our programme, we feel our training has impact when :
• Trainees are able to undertake appropriate analysis / use resources and tools introduced during the course
• The ability to use what they have learnt has enabled / improved their ability to do research
• They have transferred the knowledge and skills gained to others
Based on learning from an earlier long-term feedback assessment of transcriptomics course trainees, we designed a brief electronic survey asking a series of 10 questions focused around our impact definition.
To date, the response rate to feedback request across the first 10 courses from 2015 is 27%. The initial results provide directly applicable information for further course development and further impact assessment.
This approach may have limitations, but does provide us with a base from which we can evolve our impact measures over time.
Short Abstract: The Bioinformatics Training Platform (BTP) has been developed to provide access to the computational infrastructure required to deliver sophisticated hands-on bioinformatics training courses, without the need for dedicated training suites. The BTP is a cloud-based solution that is in active use for delivering training in topics such as next generation sequencing, metagenomics, cancer genomics and de novo genome assembly to Australian researchers at geographical dispersed locations. The BTP was built to provide an easy, accessible, consistent and cost-effective approach to delivering workshops at host universities and organisations with a high demand for bioinformatics training but lacking the dedicated bioinformatics training suites required. Being cloud-based it also minimises the amount of local IT support required to set up computer rooms for bioinformatics training. To support broad uptake of the BTP, the platform has been made compatible with multiple cloud infrastructures including the Australian research cloud and Amazon Web Services, as well as making VirtualBox and VMWare images available for trainees. The BTP is an open source and open access resource. All course materials and configuration files necessary to create the cloud virtual machine images are freely available on GitHub: https://github.com/BPA-CSIRO-Workshops. To date, over 20 training workshops have been delivered to over 700 trainees at 10 different training venues across Australia using the BTP.
Short Abstract: Specifically targeting PhD students in bioinformatics, biology and computational biology, and inspired by the success of similar initiatives in evolutionary biology, we have initiated the “Bioinformatics in the Chalet” workshops. This workshop differs from most other training events in that the participants are isolated in a high mountain chalet with a few faculty members and asked to work in pairs with a single aim: to come up with an original research project.
Throughout the week, the students are guided towards the writing process, first alone then in pairs, with daily deadlines, and with faculty mentoring all along. Every evening there is an “armchair” talk, where a faculty presents a research story without slides or preparation. These informal talks are specifically appreciated in the students’ feedback, for the insight into how science works and the opportunity to discuss differently with leading bioinformaticians.
The working environment and cosy setting provide a unique opportunity to promote networking of students and faculty from diverse areas converging to bioinformatics, and to develop students’ skills such as expressing scientific ideas, thinking critically, writing and defending a proposal. It is especially relevant for PhD students to lead them to reflect on, and immediately apply, the main criteria for a successful project in bioinformatics. The learnt skills and competencies will be useful all through their research careers.
This workshop is offered every two years as one of the SIB PhD Training Network activities. It is part of the SIB Training courses’ portfolio, which can be found at www.sib.swiss/training.
Short Abstract: GOBLET is a global organisation that coordinates, shares and supports bioinformatics training activities worldwide, aiming to plug critical skills gaps, ultimately to facilitate the advancement of health- and life-science research. The focus of GOBLET’s Train-the-Trainer initiative is on setting up effective training courses to help plug known skills gaps, especially in the area of NGS data analysis. This initiative will help to share bioinformatics training expertise, experience and resources; train bioinformatics and life-science specialists; support life-science research; promote collaborations among scientists worldwide; build capacity in developing and developed countries. The programme will consist of workshops, which will take place on different continents (e.g., South America, Africa, Asia) and are expected to co-locate as satellite events to major conferences. Each workshop is organised around two main topics: 1) how to exploit NGS data, and 2) how to set up and deliver excellent training courses. Trainers (members of GOBLET, expert in the field) will teach on a volunteer basis. Workshop participants will commit to replicate the workshops at least once, driving an exponential effect. To deliver this ambitious project, GOBLET is seeking partners and sponsors interested in increasing the provision of bioinformatics training in the area of NGS, either as GOBLET collaborators to customise the programme to specific communities or to fund workshops at given locations. To learn more about this project, see http://www.mygoblet.org/content/fund-raising, contact frc@mygoblet.org, talk to us in GOBLET’s booth and visit this poster!
| Search Posters: |