Day 5: Highlights & Recap!

And that's a wrap - 700+ Posters - 400+ Talks - 10 Tracks all on a hybrid platform! The final day took off with an engaging keynote presentation by ISCB's Innovator Award winner and ended with a plethora of dynamic talks and presentations. Again, a huge thank you to all who made this conference such a successful hybrid event, bumps and all.

Please read on for more on our final day and what sets ISMB apart!

Highlights

The final keynote address of the ISMB 2022 conference was from Dr. Nuria Lopez-Bigas, the ISCB Innovator Award winner, whose research focuses on the identification of cancer driver mutations. Tumor development follows variation and selection of Darwinian evolution. To find cancer drivers, Dr. Lopez-Bigas is working on identifying signals of positive selection as this is an effective way of uncovering cancer drivers. Signals of positive selection include genes that have mutated more frequently than background mutation rates, genes with a significant regional clustering of mutations, and genes with a bias toward high functional mutations.

The Lopez-Bigas research team is also working on a compendium of mutational drivers which aims to identify as yet unknown cancer driver genes. In addition, they’ve created BoostDM, inspired by evolutionary biology, which uses machine learning to uncover feature combinations that define driver mutations in each gene and tumor type. The findings have been integrated into the Cancer Genome Interpreter and the hope is to personalize cancer medicine and subsequent treatment, as well as to empower patients to better understand their data, and make them aware that their data has value and that in sharing it, they’re helping not only themselves but the research community and, more importantly, other
cancer patients.

COSI Recaps:

Bio-Ontologies COSI
This year’s ISMB marked the 25th year of the Bio-Ontologies COSI, which makes it the longest running COSI at ISMB. To celebrate, we featured outstanding keynotes from Dr. Maria Keet and Dr. Melissa Haendel, the latter of which took place during a shared session with the Bioinformatics Open Source Conference (BOSC) community. The COSI featured 18 paper (2 proceedings papers) and poster presentations and featured research on a wide range of topics including the detection and prevention of bias encoding in ontology development, evaluation of named entity recognition systems and automated solutions to identifying inconsistencies in ontology annotations, improved integration of clinical data and
biomedical ontologies to facilitate rare disease research, strategies to improve ontology FAIRness, new deep learning methods to improve and enhance ontology axioms, and we even saw the introduction of several new ontologies in the domains of cognitive-behavioral symptomatology, Sickle Cell Disease, and Rice Trait concepts.

BioVis COSI
BioVis 2022 started with a keynote by Tatiana von Landesberger (University of Cologne, Germany) on the very timely topic of visual analytics approaches for pandemic data. This included efforts on containing the ‘silent pandemic’ of antibiotic resistant pathogens and of course the ongoing COVID-19 pandemic, as well as reflections and lessons learned. The keynote was followed by one long abstract talk and four short ones. The talks in this session focused on single-cell data analysis and clustering approaches. In the lunch break, the posters were presented, leading to many lively discussions. The second session featured more abstract talks, ranging from metagenomic profile visualization to cohort analysis and the visual analysis of tumor phylogenies. The closing session featured another long abstract talk as well as lightning talks by all poster presenters and a keynote talk by Lennart Martens (Ghent University, Belgium), who shared his research and insights on new perspectives on proteomics through machine learning, especially on post-translational modifications of proteins and the visualization challenges that come with this. Overall, more than two thirds of the talks were given live from the venue. We had a very active participation not only by the on-site attendees but also from the online participants.

EvolCompGen
We enjoyed another full day of fantastic presentations and posters from researchers around the world on subjects as diverse as protein domain architecture evolution, conserved transcripts clusters databases, statistical models for protein alignment, comparative genomics applied to brain tumors, learning genetic features associated with types of diseases and infections driven by pathogens, selection mechanisms underlying convergent traits, or the use of evolution and machine learning to identify multiscale molecular features of antibiotic resistance. The afternoon track ended with a panel discussion where in-person and virtual attendees exchanged ideas on fundamental questions related to gene/synteny/domain/isoform evolution and large-scale phylogenetics. We thank all in-person and virtual attendees for their patience and participation as we conclude our 2022 EvolCompGen program! Stay tuned for announcements on the (non-proceedings) talk and poster winners. We also welcome you to join our COSI community and the 2023 ISMB EvolCompGen planning/organizing committee via Twitter: @EvolComp, Web: evolcompgen.org, or Slack: bit.ly/join_evolcompgen.

We look forward to seeing you in 2023!

Special Sessions:

COVID-19
The special session on COVID-19 research brought us some truly unique findings! First was the unexpected discovery from the Host Genetic Initiative that the 3p21.31 locus has been found to be associated with severity of and complications associated with COVID-19. HGI also hopes that by working together as a community we can uncover more real-time, reliable results and give us a better picture of what is happening with the spread of COVID-19 and its infection profile.

Next was the investigation by Cyclica Canada into off-target drug repurposing who found a rational drug-centric coronavirus repurposing candidate, capmatinib, which showed efficacy against multiple coronaviruses. The explanation for which was that its antiviral activity is not driven by its primary target.

Finally, we heard about the discoveries from the open science Serratus team, who are hoping to be able to predict and prevent another pandemic. The team analyzed 5.7 million sequencing libraries (in 11 days!), and have discovered 9 new species of coronavirus, all in aquatic organisms. The project has helped to show the importance of reanalyzing data, especially since we can now access global levels of data. Reanalysis will help us to create meaningful knowledge from older data that we can distil down to human wisdom that can have large impacts on human response to viruses like COVID-19. Serratus exemplifies what it means to be at the cutting edge of computational virology, and hopes that computational virologists continue to be drivers to COVID-19 research.

Equity-Focused Research
During the Equity-Focused research presentations, we first heard a discussion on race versus genetics in clinical decision-making, reporting that non-white populations are heavily underrepresented in UK Biobanks. In addition to requiring greater ethnic representation in the UK Biobank, there was a clear message that we need to let genetics talk to us instead of making assumptions based on how we as researchers define someone else’s race.

There was also a discussion of how some genome editing tools may be qualified as non-stochastic, which can be problematic. Specifically, if some genome editing and assessment tools are programmed to target genetic regions that are much more common in some populations over others, conclusions about health, treatment, and potential outcomes may be harmfully incorrect.

A third talk during the equity session presented the idea that in the field of biodiversity genomics, it would be in the best interest of science for there to be a low-cost DNA sequencing  method which would allow people beyond academia to become part of and engage in genomic sequencing, regardless of economic station. The Amazon Biobank, a community-based genetic database, has been offered as a solution to increase access to biodiversity research. The project combines blockchain and smart contract technologies to promote adequate sharing among all those who collect, insert, store, and process genomic data.

Finally was a talk about bias in science journalism. From an analysis of 22, 000 non-research articles published by Nature, two sources of disparity were identified: gender and perceived regional origin of a name. The results uncovered an obvious skew towards quoting and citing males in Nature science journalism-related articles, but the trend is moving toward equal representation more quickly than what is found in academic publishing. With regard to an analysis of name origin, there was a significant over-representation of names with predicted Celtic/English origin and an under-representation of names with a perceived East Asian origin. It is clear that greater representation of both gender and ethnicity is necessary.

These talks are now available on ISMB On-Demand.

2022 Conference Support Fellowships
Each year, conference organizers are committed to providing support to as many eligible applicants as possible. Of the 155 applications received, ISCB was able to grant 47 Awardees up to 50% of their costs associated with conference attendance, both virtually and in-person. This years group of Fellowships represented 6 continents and 17 countries.

Conference Support Fellowship consideration is based on membership and accepted work to Conferences.

Thank you to our generous sponsors!

See you next year in Lyon, France!