In Silico Testing of Hypotheses for the Effect of Smoking on Somatic Evolution in the Healthy Human Lung

Principal Investigator

Name
Sam Janes

Degrees
M.D., Ph.D.

Institution
University College London

Position Title
Director, Division of Medicine

Email
s.janes@ucl.ac.uk

About this CDAS Project

Study
PLCO (Learn more about this study)

Project ID
PLCO-1258

Initial CDAS Request Approval
Jun 22, 2023

Title
In Silico Testing of Hypotheses for the Effect of Smoking on Somatic Evolution in the Healthy Human Lung

Summary
Recent single-cell genomic analysis of healthy lung tissue has shown remarkable intra-tissue heterogeneity in the degree of effect smoking has on mutational burden, as well as an expansion of less-mutated basal cell sub-populations after smoking cessation. These two findings suggest potential mechanisms for somatic evolution in the healthy lung; here, we use computational modelling, based on a model of lung homeostasis previously verified by lineage tracing, to assess the ability of these hypotheses to reproduce observations. Applying a Bayesian inference framework to simulations of basal lung cell populations over the course of patients’ lifetimes, we find evidence for a protected sub-population of basal cells in the lung which are less affected by smoking. First-principles modelling of underlying mechanisms to compare hypotheses provides insight into the way cancer develops in the lungs. This could allow for more targeted analyses of relevant datasets to improve risk modelling and streamline the search for novel preventative treatments.
In order to validate the computational model of somatic evolutionary processes in the healthy human lung, we intend to compare the model with epidemiological datasets of cancer risk stratified by smoking status.

Aims

1. Define and implement computationally a model of somatic evolution in the normal airway epithelium, including separate modules for each modelling hypothesis and a framework for fitting the models’ parameters to the data and assessing the goodness-of-fit.
2. Assess all aspects of the model for robustness and accuracy, to ensure any inference can be relied upon.
3. Assess outputs of the system created in Aim 1 to deepen understanding of the mechanisms, and check incidental predictions of the simulations against external datasets of cancer epidemiological data.

Collaborators

Kate Gowers, UCL
Vitor Teixeira, UCL
Calum Gabbutt, Institute of Cancer Research
Carlos Martinez-Ruiz, UCL
Ahmed Alhendi, UCL
Ben Simons, University of Cambridge
Sam Janes, UCL
Nicholas McGranahan, UCL
Adam Pennycuick, UCL