This is the first newsletter of 2020, already two and a half years into our second round of funding! The year has already been busy as we are preparing our training program and getting projects back on track after the break. It was difficult to schedule all our planned online courses into the calendar year as we have 3 ongoing ones (IBT, IntBT 16S rRNA data analysis and Genomic Medicine) and two new ones: Train the trainer and IntBT NGS. In order to avoid too much overlap as some sites only have 1 available classroom, we have had to postpone the launch of the IntBT NGS to early 2021. Our training and outreach team has worked with the H3Africa Coordinating Centre to run a professional development work in Cape Town (see report in this newsletter), and together we are planning a DNA Day event with the NHGRI and providing access to an online health Informatics course.
As part of capacity building and resource development, our machine learning project is developing a handbook of Machine Learning Tools, Platforms and Packages for Bioinformaticians. See below for an article on this activity, we hope you find them useful if you are just venturing into this field. I am pleased that our Helpdesk is back up and running. We had some down time as we upgraded to a new backend system. This took a few rounds of installation and testing as we sought the best system to use that had the required functionality without a high price. The final article describes the new features and how the helpdesk functions. I hope you enjoy reading this and the rest of the newsletter.
A Participant’s Perspective on the H3Africa Professional Development Programme hosted in Cape Town during 2019
By: Michelle Namuyaba
My name is Michelle Namuyaba. I am a current post graduate student studying a Master of Science in Bioinformatics with Systems Biology at Birkbeck College, University of London. My journey with Bioinformatics began in 2018. It was my final year doing my undergraduate degree in Biomedical Laboratory Technology at Makerere University in Uganda. One of my course modules was Bioinformatics being taught by Dr. Gerald Mboowa. Dr.Gerald Mboowa taught the subject with so much passion, the content he spoke about this field intrigued me so. At the end of my course, a colleague of mine by the name of Maria Namaganda that was interested in the field as well shared with me the call for participants to be a part of the H3AbioNet 3 month Introductory course to Bioinformatics (IBT). I was able to apply and I was successfully chosen to be among the individuals to part take in this course. I learnt a great deal from this course. I was also able to be further informed about the organisation that is H3Africa that oversees the entire program. I learnt that not only does H3Africa offer online courses to scientists but also holds multiple annual conferences for scientists across the continent to share their content and learn from one another to better the scientific community in Africa. This information, alongside the motivational words of my mentor Dr. Gerald Mboowa motivated me to apply for the various calls to be a part of the conferences hosted by H3Africa in 2019. I was successfully chosen to attend the Professional Development Workshop hosted by H3Africa in Cape Town South Africa 2019. This invitation came along with a travel award. I was so overwhelmed and excited to be a part of this community. On attendance of this conference, I learned that women in the scientific community can indeed build and encourage each other in a society that traditionally had a larger percentage of male individuals and promote equality. The workshop showed me how far teamwork, clear and concise communication and collaboration can take you. I learnt great presentation skills that I was able to use once I got back to my home country, as well as during the course of my postgraduate study. I was also able to meet various accomplished scientists from around the continent. This enabled me to improve my networking skills. Attending the conference drove me to make a clearer and more focused plan for my future as a female scientist. As a result, I was able to apply for and be chosen to study Bioinformatics at Birkbeck College in London, United Kingdom. I am forever grateful to my mentor for always encouraging me and the H3Africa team for giving me the opportunity to be a part of such a community: as I believe they collectively paved a way for me to make career plans and chase them regardless of my situation.
Machine Learning has been around for a few years now and this has resulted in a compendium of machine learning tools, packages and platforms with new ones being created continuously. A bioinformatician wishing to use machine learning will need to spend much time looking into the different existing possibilities before being able to choose the correct tool/platform/package. A number of resources and handbooks on machine learning tools are available in the public domain but the majority of them are intended for users with computing backgrounds. The H3ABioNet Handbook of Machine Learning Tools, Platforms and Packages for Bioinformaticians/Directory of Machine Learning Tools, Platforms and Packages provides an overview and guidelines on the use of some of the most popular machine learning tools, platforms and packages that could be used for common bioinformatics and genomics analysis tasks.
The current version of this handbook provides detailed descriptions for over 40 tools, platforms and packages. For each tool/platform/package, the following information has been provided:
- whether it is open source or not – this will guide the user in the budget planning
- where it can be accessed or downloaded
- the expected input and output data types – the input can be text, audio, video, image or specific formats such as csv, sql, etc. The output format is important since it will impact on how the output will be further processed or saved
- the operating system/s on which it can be used – this is an important consideration since some of the tools or packages are specific to some operating systems while others can only be accessed via Cloud
- the limitations, if any – this will guide the user in the final choice of tool depending on what the user expects from the tool
- a small description – for users to know more about the tool or package or platform
- and its use in bioinformatics and genomics – some tools have already been used for very specific tasks such as genomics data analysis, sequence analysis, protein structure prediction, etc while others have been used for classification, regression etc in other fields and can potentially be used for similar tasks in bioinformatics.
H3ABioNet Pilots the use of Containers for Advanced Bioinformatics Training in Africa
By: Verena Ras
From October to November 2019, H3ABioNet made the brave decision to pilot a brand new 16S rRNA microbiome data analysis course, using the popular blended learning model that has made our introduction to bioinformatics training course so popular over the last few years. With data increasingly being generated on the African continent and with so little data analysis support, the overarching aim of this project was to build local capacity in this type of analysis.
The development of the course was two-fold. It involved detailed curriculum development alongside expert trainers who performed thorough competency mapping. This led to the development of 6 core modules, covering topics such as study design, quality control, ASV picking, nextflow workflow management, Unix and R. It also had a strong technical component, where the technical team was tasked to find a method allowing all classrooms access to specific reference datasets but to also ensure they have the tools/software readily available to analyse the data - and with sufficient computational power.
But with big data comes big computational demands….
This meant the core team had a big decision to make. They either managed all infrastructure locally or, they dared to do what was thought to be a mammoth task… to have local hosting sites manage infrastructure independently with some support from the core team. Since the goal was to develop local capacity around Africa, they chose the more difficult of the two paths and requested sites pull developed containers.
But this had one major catch… If it was to be run this way, unlike IBT, hosting sites needed to pull training containers, manage and maintain them independently (with some remote support from H3ABioNet) and this required quite a bit of expertise and meant the core team had hundreds of queries which they managed with finesse.
Although there were many doubts, the course ran successfully. The first iteration had 23 hosting sites formally accepted which, subsequently enrolled around 250 participants across these sites. The containers appeared to work well and had the added benefit of the infrastructure being available for as long as it is maintained. This meant sites could analyse data well after the course concluded… one of the major goals of the project.
We hope to run the second iteration in 2020 (September 2020 TBD) and with new developments and improvements you do not want to miss out so keep an eye on our website for course calls and announcements!
During the first round of H3ABioNet, the network established one of the first public and freely available, online bioinformatics helpdesks to support the development of bioinformatics and genomics capacity in Africa and provide rapid bioinformatics support to researchers across Africa and the rest of the world. Since then, the Helpdesk has been under development, and here we introduce the new and improved H3ABioNet Helpdesk.
Our Helpdesk provides support through Helpdesk representatives, with backgrounds and significant experience in various bioinformatics applications. Representatives are voluntary members of H3ABioNet, and range from mid- to late-career scientists, with unique experience in bioinformatics application in Africa. A Helpdesk administrator prioritizes queries and assigns them to the appropriate Helpdesk representative, based on the overlap between a representative’s expertise and the query category.
Our Helpdesk accepts queries, submitted as tickets, through an online interface accessible from the H3ABioNet website - https://helpdesk.h3abionet.org/. No logins or accounts are needed, and the Helpdesk is open to all who require assistance (see Figure on the left).
We support a very wide range of bioinformatics queries, including:
- Data Analysis help for NGS, 16S and GWAS Data,
- Various pipeline assistance such as Imputation and Variant Calling
- System/Technical Administration assistance
- Data Management and Biostatistics assistance,
- Assistance using the H3Africa GWAS Chip and Standard CRF,
- Assistance using developed tools, such as the Human Mutation Analysis (HUMA) platform, REDCap, NetCapDB and Minimum Data Dictionaries,
- More info on our Training Courses, Website and Social Media,
- And more!!!
- On our new platform, you are allowed to assign a priority and due date to your submitted ticket, this allows us to prioritize your ticket and get back to you as soon as possible! Moreover, you can also add attachments to your queries - this is useful when you’re running into specific errors during your data analysis and you need to communicate this to the Helpdesk representatives. Finally, we can also track the assistance we provide, so we know which H3Africa projects need more assistance, and which training is still needed across the continent! See Figure below.
Our Helpdesk’s performance is regularly monitored and evaluated, and we very much encourage user feedback! Previous performance analysis has revealed that the Helpdesk performs efficiently and resolves queries in a timely manner. User feedback has been equally positive. The majority of Helpdesk users rated quality of support (77%), query resolution time (74%), and user friendliness of the Helpdesk interface (84%), above average, on a standard 1 – 5 scale. Similarly, 94% of these users also revealed that they would use the Helpdesk again if needed in the future and recommend use of the Helpdesk to other users.
Students, researchers and bioinformaticians are all encouraged to use the H3ABioNet Helpdesk to gain project-specific support for the issues they may be facing with their analyses and (or) research! The H3ABioNet Helpdesk is actively promoted through outreach material and at H3ABioNet training events! We are happy and eager to assist you wherever we can! If you would like to get in touch with us or have any further questions, please submit them to the Helpdesk (General Enquiries) or contact Lyndon Zass (email@example.com) or Mamana Mbiyavanga (firstname.lastname@example.org).