A hybrid modeling approach for assessing mechanistic models of small molecule partitioning in vivo using a machine learning-integrated modeling platform

Abstract

Prediction of the first-in-human dosing regimens is a critical step in drug development and requires accurate quantitation of drug distribution. Traditional in vivo studies used to characterize clinical candidate's volume of distribution are error-prone, time- and cost-intensive and lack reproducibility in clinical settings. The paper demonstrates how a computational platform integrating machine learning optimization with mechanistic modeling can be used to simulate compound plasma concentration profile and predict tissue-plasma partition coefficients with high accuracy by varying the lipophilicity descriptor logP. The approach applied to chemically diverse small molecules resulted in comparable geometric mean fold-errors of 1.50 and 1.63 in pharmacokinetic outputs for direct tissue:plasma partition and hybrid logP optimization, with the latter enabling prediction of tissue permeation that can be used to guide toxicity and efficacy dosing in human subjects. The optimization simulations required to achieve these results were parallelized on the AWS cloud and generated outputs in under 5 h. Accuracy, speed, and scalability of the framework indicate that it can be used to assess the relevance of other mechanistic relationships implicated in pharmacokinetic-pharmacodynamic phenomena with a lower risk of overfitting datasets and generate large database of physiologically-relevant drug disposition for further integration with machine learning models.

Figure 1

Basic diagram of quasi-2D tissue distribution in a compartmental model. D_tissue is the diffusion coefficient for a compound across the plasma-tissue barrier; fu refers to unbound drug concentration in plasma (p) and tissue (t), and Q refers to blood flow rates through a tissue section.

Figure 2

logP optimization convergence plots for 21 small molecule compounds tested with logP optimization. Y-axis corresponds to the cost function; X-axis is the cost progression across grid search optimization (high oscillation) and descent (minimal oscillation) simulation trials.

Figure 3

Plots of Observed vs. Predicted PK metrics across the different optimization conditions for firstt-order PK outputs. Red lines correspond to lines of best fit, gray lines are bounds of ± threefold-error.

Figure 4

Plots of Observed vs. Predicted PK metrics across the different optimization conditions for second-order PK outputs. Red lines correspond to lines of best fit, gray lines are bounds of ± threefold-error.

Figure 5

Comparison of the PK outputs fold-difference magnitude and geometric mean fold-error (GMFE) across the three optimization conditions.

Figure 6

Comparison of the simulated Kp values to the Rodgers-calculated values for gut and muscle compartments with linear regression best-fit.

Figure 7

Overview diagram of the BIOiSIM mechanistic model. Note: some compartments are omitted for clarity. CL_liver hepatic clearance, Cl_renal renal clearance.