Incorporating uncertainty to improve accuracy in mathematical modeling of coagulation
Abstract
Hemostasis is the normal process by which the body stops bleeding in the event of an injury. Blood coagulation is an important part of hemostasis where dozens of proteins act collectively to initiate a rapid clotting response. Ineffective coagulation leads to bleeding; excessive coagulation leads to pathological clot formation. Components of the coagulation system can thus serve as targets for therapeutic agents designed to be either prohemostatic or anticoagulant. However, there are currently no experimental assays capable of identifying individual outcomes of such therapeutics; variability amongst individuals, noise in experiments, and the multiple positive and negative feedback loops inherent to coagulation make its response challenging to predict. Accurate mathematical models of coagulation would complement existing experimental assays in risk prediction and development of therapeutics. A few mathematical models have been developed to describe coagulation; however, despite years of effort the comparison between such models and experiments is typically only qualitative. Further, the behavior captured by such mathematical formulations is not robust under experimental manipulations of specific chromogenic substrates, inhibitors, lipids, and other factors. The overall goal of the proposed work is to establish and experimentally verify a unifying mathematical model of coagulation that can accurately account for thrombin generation and prothrombin conversion. Development of this model is a critical step in creating a highly predictive tool with potential to aid in the design of new prohemostatics and anticoagulants. In contrast to prior modeling efforts, our innovative approach will assess the uncertainty inherent in an experimental system by including: 1) experimental noise, 2) error or uncertainty in initial conditions, 3) error or uncertainty in kinetic rate constants, and 4) the possibility of incorrect or incomplete kinetic schemes. To achieve our goal, we will focus on three specific aims: (1) develop and validate models of chromogenic substrate to determine how they alter the biochemistry of the system; (2) infer kinetic rates for lipid-surface dependent coagulation reactions, and (3) identify the mechanisms that regulate prothrombin conversion to thrombin. Our model is expected to have a positive impact; it will seamlessly and accurately probe the coagulation network and thus provide a more detailed analysis of the pathways and rate constants leading to thrombin generation. It could also be used to inform anticoagulant treatment strategies and speed the development of prohemostatic agents by helping to find optimal biochemical targets or combinations of synergistic targets that treat or prevent bleeding.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Oct 11, 2018
- Source ID
- W911NF1710465
Entities
People
- Suzanne Sindi
Organizations
- Army Contracting Command
- United States Army
- University of California