Multiscale simulation of thrombus growth and vessel occlusion triggered by collagen/tissue factor using a data-driven model of combinatorial platelet signalling.

During clotting under flow, platelets bind and activate on collagen and release autocrinic factors such as ADP and thromboxane, while tissue factor (TF) on the damaged wall leads to localized thrombin generation. Towards patient-specific simulation of thrombosis, a multiscale approach was developed to account for: platelet signalling [neural network (NN) trained by pairwise agonist scanning (PAS), PAS-NN], platelet positions (lattice kinetic Monte Carlo, LKMC), wall-generated thrombin and platelet-released ADP/thromboxane convection-diffusion (partial differential equation, PDE) and flow over a growing clot (lattice Boltzmann). LKMC included shear-driven platelet aggregate restructuring. The PDEs for thrombin, ADP and thromboxane were solved by finite element method using cell activation-driven adaptive triangular meshing. At all times, intracellular calcium was known for each platelet by PAS-NN in response to its unique exposure to local collagen, ADP, thromboxane and thrombin. When compared with microfluidic experiments of human blood clotting on collagen/TF driven by constant pressure drop, the model accurately predicted clot morphology and growth with time. In experiments and simulations at TF at 0.1 and 10 molecule-TF/$\mu$m$^{2}$ and initial wall shear rate of 200 s$^{-1}$, the occlusive blockade of flow for a 60-$\mu$m channel occurred relatively abruptly at 600 and 400 s, respectively (with no occlusion at zero TF). Prior to occlusion, intrathrombus concentrations reached 50 nM thrombin, ~ 1 $\mu$M thromboxane and ~ 10 $\mu$M ADP, while the wall shear rate on the rough clot peaked at ~ 1000-2000 s$^{-1}$. Additionally, clotting on TF/collagen was accurately simulated for modulators of platelet cyclooxygenase-1, P2Y$_{1}$ and IP-receptor. This multiscale approach facilitates patient-specific simulation of thrombosis under hemodynamic and pharmacological conditions.

A human platelet calcium calculator trained by pairwise agonist scanning.

Since platelet intracellular calcium mobilization [Ca(t)]i controls granule release, cyclooxygenase-1 and integrin activation, and phosphatidylserine exposure, blood clotting simulations require prediction of platelet [Ca(t)]i in response to combinatorial agonists. Pairwise Agonist Scanning (PAS) deployed all single and pairwise combinations of six agonists (ADP, convulxin, thrombin, U46619, iloprost and GSNO used at 0.1, 1, and 10xEC50; 154 conditions including a null condition) to stimulate platelet P2Y1/P2Y12 GPVI, PAR1/PAR4, TP, IP receptors, and guanylate cyclase, respectively, in Factor Xa-inhibited (250 nM apixaban), diluted platelet rich plasma that had been loaded with the calcium dye Fluo-4 NW. PAS of 10 healthy donors provided [Ca(t)]i data for training 10 neural networks (NN, 2-layer/12-nodes) per donor. Trinary stimulations were then conducted at all 0.1x and 1xEC50 doses (160 conditions) as was a sampling of 45 higher ordered combinations (four to six agonists). The NN-ensemble average was a calcium calculator that accurately predicted [Ca (t)]i beyond the single and binary training set for trinary stimulations (R = 0.924). The 160 trinary synergy scores, a normalized metric of signaling crosstalk, were also well predicted (R = 0.850) as were the calcium dynamics (R = 0.871) and high-dimensional synergy scores (R = 0.695) for the 45 higher ordered conditions. The calculator even predicted sequential addition experiments (n = 54 conditions, R = 0.921). NN-ensemble is a fast calcium calculator, ideal for multiscale clotting simulations that include spatiotemporal concentrations of ADP, collagen, thrombin, thromboxane, prostacyclin, and nitric oxide.