Conference Summary: AIRR Community Meeting VII – Learnings and Perspectives

At the start of June, we (Lewis and Benjie) attended the AIRR Community meeting in beautiful and sunny Porto, Portugal. This meeting was focused on collecting and analysing adaptive immune receptor repertoires. This comprised of two rivalling factions at the conference: the antibody (Ab) people or the T cell antigen receptor (TCR) people. The split was nearly fifty-fifty between these two topics throughout the conference. Overall, the conference was a fairly comfortable size, with approximately a hundred people in attendance, making it easy to visit all of the posters and talk with many people in your area, without feeling too niche. There was a wide variety of content formats throughout the conference including posters, scientific talks, lightning talks, software demos, and hands-on tutorials. In the following section, we highlight some of our favourite sessions to give a flavour of what this meeting entails.

Leveraging single-cell TCR-sequencing to investigate mucosal-associated invariant T-cell biology

Nicholas Provine of Oxford University gave an excellent talk on the single-cell analysis of MAIT cells coming from multiple donors, looking at the similarities and differences of these cells between people. For those of you wondering what the heck a MAIT cell is, Nicholas’ talk gave a great view of the role of these cells in our immune systems. MAIT cells, standing for Mucosal Associated Invariant T cells, are a hyper-abundant type of T cell that are found as their name suggests, mainly in mucosal sites. What makes these cells special is that they have a fixed alpha chain of their TCRs and they recognize small molecule antigens, such as riboflavin, presented by MHC-like MR1 molecules.

The talk highlighted the broad specificity of these cells and interestingly, found that although their alpha chain is fixed, there is a lot of diversity in the beta chain pairings that make them “private” in the donor repertoires. The following figure highlights this, showing that the clonotypes are not found across different donors as a result of the diverse beta chains.

The talk also discussed the differences in MAIT cell functions after being activated by TCR signalling or by cytokine stimulation, and that these activation pathways lead to different downstream functions in the cells. The work can be read about more in the following Nature Immunology article.

Decoding the adaptive immune system using deep generative models

On the computational modelling side, Niranjani Prasad gave a talk about Adaptive Biotechnologies powerful data pipeline and AIRIVA generative modelling framework.

The talk showed the impressive amount of TCR sequencing data Adaptive has collected from a multitude of disease contexts and patient cohorts. Niranjani showed their statistical analysis framework for associating TCRs with a given disease context, showing the applicability of using these identified TCRs as biomarkers for the disease. The talk then went into the generative modelling work being done at adaptive using variational auto-encoders with context labels that they call ARIVA. These models show promise in being able to learn rich representations of TCR sequences with applications in antigen specificity, therapeutic generation, and other tasks. Working in this area ourselves, it is impressive to see the scale at which a company can operate and perform research.

LZGraphs

A favourite from the lightning talks, Thomas Konstantinovsky gave a demo on his approach to encoding adaptive immune receptors as graphs. These graphs represent repertoires as directed graphs, grouping common motifs in the receptors as nodes to enable data compression and insights through the graph’s topology. We found this to be an elegant way of representing AIRR data and hope to incorporate it into our research. A pre-print on the topic can be found here.

Single-cell based antigen-receptor gene and function analyses to instruct vaccine designs

Hedda Wardemann provided interesting insights into Memory B-cell (MBC) population changes in response to SARS-CoV-2 vaccinations.

In the study, MBCs were split into two populations – those with few somatic hypermutations (low-SHM) and those with many SHMs (high-SHM) that sat further from the germline. Pre-immunisation, SARS-CoV-2 spike-specific (S+) MBCs with many SHMs existed in patients due to pre-exposure to other coronaviruses. These high-SHM MBCs formed the majority of patients’ MBC populations up to 3 weeks post immunisation (I+3w). However, despite being relatively weak binders, these MBCs did not undergo affinity maturation and instead progressed directly to become antibody-secreting cells (ASCs).

The low-SHM MBC population on the other hand showed strong signs of undergoing affinity maturation before differentiating to ASCs. This maturation meant low-SHM MBCs became the most prevalent (>75%) and high-affinity S+ population 3 weeks post the second round of immunisation (II+3w). Prof. Wardemann’s work shows vaccinations or previous infections may offer some immunity to novel disease variants, but that affinity maturation is still required for a robust immune response. Furthermore, it shows an antibody’s distance from the germline cannot be reliably used as a proxy for antigen binding affinity. More can be found in the Pre-print!