Complete populations of virtual patients for in silico clinical trials
Stefano Sinisi, Vadim Alimguzhin, Toni Mancini, Enrico Tronci, Brigitte Leeners. Bioinformatics, 36(22-23), pages 5465–5472, 2020.
Stefano Sinisi, Vadim Alimguzhin, Toni Mancini, Enrico Tronci, Brigitte Leeners.
Complete populations of virtual patients for in silico clinical trials.
Bioinformatics, 36(22-23), pages 5465–5472, 2020.
DOI: 10.1093/bioinformatics/btaa1026BibTex entry
@ARTICLE{sinisi-etal:2020:bioinf, TITLE = {Complete populations of virtual patients for in silico clinical trials}, AUTHOR = {Sinisi, S. and Alimguzhin, V. and Mancini, T. and Tronci, E. and Leeners, B.}, DOI = {10.1093/bioinformatics/btaa1026}, ISSN = {1367-4803}, JOURNAL = {Bioinformatics}, NUMBER = {22--23}, PAGES = {5465--5472}, PUBLISHER = {Oxford University Press}, VOLUME = {36}, YEAR = {2020} }
Model-based approaches to safety and efficacy assessment of pharmacological drugs, treatment strategies, or medical devices (In Silico Clinical Trials, ISCT) aim to decrease time and cost for the needed experimentations, reduce animal and human testing, and enable precision medicine.
Unfortunately, in presence of non-identifiable models (e.g., reaction networks), parameter estimation is not enough to generate complete populations of Virtual Patients (VPs), i.e., populations guaranteed to show the entire spectrum of model behaviours (phenotypes), thus ensuring representativeness of the trial. Results.
We developed methods and software based on global search driven by statistical model checking that, starting from a (non-identifiable) quantitative model of the human physiology (plus drugs PK/PD) and suitable biological and medical knowledge elicited from experts, compute a population of VPs whose behaviours are representative of the whole spectrum of phenotypes entailed by the model (completeness) and pairwise distinguishable according to user-provided criteria. This enables full granularity control on the size of the population to employ in an ISCT, guaranteeing representativeness while avoiding over- representation of behaviours.

Results
- Effectiveness of our algorithm proved on a non-identifiable ODE-based model of the female Hypothalamic-Pituitary-Gonadal axis
- Computed a population of 4 830 264 virtual patients
- Virtual patients stratified into 7 levels at different granularity of behaviours
- Completeness of population assessed against 86 retrospective health records from Pfizer, Hannover Medical School and University Hospital of Lausanne: datasets covered by our VPs within Average Normalised Mean Absolute Error of 15%, 20%, and 35%, respectively (90% of the latter dataset is covered within 20% error).