Data mining of human clinical data for validation of mathematical model of host-pathogen interactions in lungs due to M. tuberculosis with different drug-sensitivity /Анализ клинических данных для верификации математической модели взаимодействия в системе хозяин-патоген, обусловленного штаммами M. tuberculosis с различной лекарственной чувствительностью

Project: Grant fulfilmentTotal grant fulfilment

Project Details

Description

The Mycobacterium tuberculosis (Mbt)- causative agent of tuberculosis- manipulates the host immune response and thus stimulates inflammatory processes leading to an imbalance of matrix metalloproteinases and its inhibitors (MMP/inhibitors) facilitating tissue destruction. Moreover, different resistant strains might have functional dynamics reflected in changes of MMP/inhibitors concentrations, distinct from the dynamics of drug-sensitive strains that can lead to different outcomes of the planned treatment. The association of MMP/inhibitors imbalance and various clinical forms of lung tissue destruction depending on the drug sensitivity of the pathogen was not investigated before. The proposed study aims to fill this gap. In particular, we will leverage advanced multidimensional statistical methods applied to clinical outcome data to validate our mathematical model and to assess the potential of novel biomarkers.

The goals of the collaborative project are:
- to conduct a large-scale multidimensional statistical analysis that takes into account many clinical parameters, such as the concentration of major metalloproteinases, the degree of bacterial excretion in patients, the number of destruction foci, the degree of tissue destruction in each focus (based on X-Ray images). Major statistical tools will include multidimensional analysis, random forest regression and classification methods as well as survival analysis.
- to validate our mathematical model of the host-pathogen dynamics in lungs, incorporating manipulation of metabolic pathways and immune response by the pathogen, taking into account the data collected at SPb Institute of Phthisiopulmonology and from the literature.
- to carry out a comparative analysis of the model obtained with the existing models on host-pathogen dynamics [Holling, 1959); Bazykin A.D, 1998], in order to obtain key parameters that control the spread of the pathogen in the lungs to predict the initial stages of treatment.
- to explore data obtained by the methods of Raman spectroscopy , to differentiate between drug-resistant and drug-sensitive pathogens, as well as of pathogen genomic data (collected at the SPSU) as possible additional covariates of treatment outcomes.
- to prepare a scientific publication and to write a grant proposal for e.g.Volkswagen Stiftung to ensure the further collaboration.

Key findings for the project

проведен статистический анализ имеющихся клинических данных для выявления основных предикторов / показателей клинических результатов лечения. В частности, были использованы передовые многомерные статистические методы, применяемые к данным клиническим результатам, для валидации математической модели и оценки потенциала новых биомаркеров.
- параметры математической модели достоверно оценены с использованием результатов статистического анализа.

Планируемые публикации:
Plos Computational Biology IF 3.955 or/and Physical Review E, IF 2.366

Дальнейшие этапы реализации проекта (вне рамок программы поддержки совместных проектов СПбГУ и СУБ):
Для продолжения сотрудничества по данному проекту планируется написание заявки на грант от фонда Volkswagen Stiftung.
Short titleMathematical modelling of Mtb population dynamics in lungs
AcronymJSMF 2019
StatusFinished
Effective start/end date15/06/1931/08/19

Keywords

  • M.tuberculosis
  • clinical data
  • statistical analysis
  • mathematical model
  • pathogen-host dynamics