Network webinars

The first QuanTII webinar took place on 20 May 2020 - it was delivered by Paolo Vicini from Kymab (P7), and was well attended by network members.

Title: Systems Pharmacology and Pharmacometrics: How Mathematics, Statistics and Data Science Are Impacting Drug Discovery and Development

Abstract: Modern drug discovery and development are rapidly becoming more reliant on rigorously quantitative approaches. In addition to established statistical testing and experimental design techniques, new approaches include pharmacometrics and systems pharmacology. Pharmacometrics is a collection of quantitative tools applied to clinical drug development and trial design, including mixed effect models for drug exposure and response, and covariate (explanatory variables) selection methods to quantify the impact of demographic, disease status and genetic variation on drug dosing, concentration and effect. Systems pharmacology is an evolution of systems biology, and seeks to harness quantitative, time-dependent pathway models to predict and quantify the effects of pharmacological interventions on downstream biomarkers, ultimately aiding rational target and drug candidate selection. This presentation will describe modern drug discovery and development pipelines and the role of pharmacometrics and systems pharmacology, highlighting in particular their interdisciplinary nature and the extent to which they borrow from other disciplines, including mathematics, statistics and computer science.

The second webinar was given by Léa Sta (ESR 3) and was attended by all ESRs, three partners, as well as external participants consisting of MSc/PhD students at the University of Leeds and Imperial College London, and an academic from Comillas Pontifical University.

Title: An analytical study of a IL-7R model

Abstract: IL-7 is a cytokine necessary for the survival of T cells. Its receptor (IL-7R) is composed of the common gamma chain and a specific alpha chain. When signalling, the IL-7R uses the JAK-STAT pathway, thus the intracellular molecule JAK3 is necessary for IL-7 signalling. Gregoire Altan-Bonnet's team, at NIH, made the following counter-intuitive observation: when the number of gamma chains on the surface of a cell increases, IL-7 signalling decreases. They also observed that an increasing number of gamma chains per cell increases the EC50 while the amplitude of the dose response first increases then decreases. We developed a mathematical model which explains the drop in IL-7 responsiveness by the formation of dummy complexes when the number of JAK3 molecules is limiting, and whose its amplitude behaves like in the experiment. We make use of an algebraic tool known as the Groebner basis to derive analytical expressions for the steady states of our model. These expressions allow us to make a full mathematical study and to compute asymptotic expressions. In particular, we can compute analytically the amplitude and the EC50 of the dose-response, providing insights that may be relevant for biologists interested in cytokine signalling. Our methods can be extended to other dimeric receptor models.

ESR 6, Giulia Belluccini, gave the fourth QuanTII webinar on 10 September at 11am UK time, via MS Teams.
It was attended by ESRs and partners, as well as an academic and PhD students from the University of Leeds.

Title: Multi-compartmental models for T cell proliferation and death

Abstract: Many biological processes are modelled using Markov chains, therefore the inter-event times, i.e. the times between consecutive events, are assumed to be exponentially distributed. This hypothesis fails when cell proliferation plays a key role. Indeed, the history-dependence nature of the cell cycle breaks the Markov property.
In this talk I will present a mathematical model for the cell cycle which guarantees the validity of the Markov property and includes cell proliferation and death.  The idea behind is to divide the cell cycle into several compartments, each of whom is exponentially distributed and independent on the others.
Then, cell generations will be introduced in the model to infer the parameters with CFSE data, making use of Bayesian methods.
After that, I will consider an alternative scenario where the cell decides its fate at birth.

Ken Duffy from Maynooth University gave a network webinar on 17 November 2020 at 2pm GMT, via MS Teams.
It was attended by ESRs and partners, as well as external PhD students.

Speaker: Ken Duffy, Hamilton Institute, Maynooth University

Title: Optimizing Spatial RNA Profiling of Single Cells

Abstract:
Quantitative experimental technologies are experiencing rapid advances throughout the biological sciences. One recently introduced technique experiencing that accelerated development is multiplexed fluorescent in situ hybridization (mFISH), which enables the identification and localization of large numbers of individual strands of RNA within single cells. Core to the technology is a problem in error correcting codes: with each RNA sequence of interest being a codeword, how should a codebook of probes be designed, and how to decode the resulting noisy fluorescence measurements? Assuming no prior knowledge of information or coding theory, in this talk we will explain how more accurate inferences can be extracted from existing experiments, and indicate how, by moving away from traditional coding theory assumptions towards those that are consistent with the current application, better code-books can be designed that will enable significantly more species of RNA to be captured accurately in a single experiment.

This talk is based on work with Luca D'Alessio (Broad Institute), Mehrtash Babadi (Broad Institute), Yonina Eldar (Weizmann Institute of Science), Litian Liu (MIT), and Muriel Medard (MIT).

Thomas Höfer from DKFZ gave a network webinar on 27 January 2021 at 15.00 GMT / 16.00 CET, via MS Teams.
ESRs and network members attended, as well as external participants.

Speaker: Thomas Höfer, DKFZ

Title: Everything is local: clonal T cell dynamics in graft-versus-host disease

Elena de Dios Panal and Erdem Şanal gave a joint webinar on 10 May 2021 at 15.00 BST. This was attended by ESRs and network members, as well as several external participants.

Speakers: Elena de Dios Panal, UMC Utrecht, and Erdem Şanal, Universiteit Utrecht

Title: T cell dynamics in a more natural mouse model

Abstract: Memory T-cells in tissues play a major role, not only in protection against re-infection but also in autoimmune disease. Because most research is based on T-cells in the circulation, i.e. cells isolated from human blood or mouse lymphoid tissues, it remains unclear how these cells maintain themselves in tissues. Are they maintained by self-renewal, by cellular longevity or by input from another source? Obtaining such insights will help the development of potential treatments for diseases like rheumatoid arthritis. Specific-pathogen-free (SPF) mice cannot be used to study the maintenance of these cells, as they mostly lack antigen-experienced T-cells in non-lymphoid tissue.
To overcome this shortcoming, we studied T-cell maintenance in wildling mice. These C57BL/6 mice born to wild mice have a natural microbiome, which shifts their naïve immune system to a more natural situation with antigen-experienced T-cells present in tissues. Using in-vivo deuterium labelling, we modelled the label incorporation dynamics of T-cells and estimated the lifespan of T-cells in the spleen, lymph-nodes, bone-marrow and lungs of wildlings. While the dynamics of memory T-cells in most tissues were comparable, we found that CD4+ memory T-cells in the lungs of wildlings are longer-lived than those in lymphoid tissues. We also found that CD4+ and CD8+ memory T-cells in lymphoid tissues of wildlings are shorter-lived than in SPF mice, while naive T-cells have similar dynamics in both models. Natural microbial exposure thus not only shapes the immune phenotype but also the dynamics of immune cells, and should be taken into account when studying the mechanisms of memory T-cell maintenance in an animal model.