A machine learning framework for computationally expensive transient models.

Transient simulations of dynamic systems, using physics-based scientific computing tools, are practically limited by availability of computational resources and power. While the promise of machine learning has been explored in a variety of scientific disciplines, its application in creation of a framework for computationally expensive transient models has not been fully explored. Here, we present an ensemble approach where one such computationally expensive tool, discrete element method, is combined with time-series forecasting via auto regressive integrated moving average and machine learning methods to simulate a complex pharmaceutical problem: development of an agitation protocol in an agitated filter dryer to ensure uniform solid bed mixing. This ensemble approach leads to a significant reduction in the computational burden, while retaining model accuracy and performance, practically rendering simulations possible. The developed machine-learning model shows good predictability and agreement with the literature, demonstrating its tremendous potential in scientific computing.

Monocyte-targeting supramolecular micellar assemblies: a molecular diagnostic tool for atherosclerosis.

Atherosclerosis is a multifactorial inflammatory disease that can progress silently for decades and result in myocardial infarction, stroke, and death. Diagnostic imaging technologies have made great strides to define the degree of atherosclerotic plaque burden through the severity of arterial stenosis. However, current technologies cannot differentiate more lethal "vulnerable plaques," and are not sensitive enough for preventive medicine. Imaging early molecular markers and quantifying the extent of disease progression continues to be a major challenge in the field. To this end, monocyte-targeting, peptide amphiphile micelles (PAMs) are engineered through the incorporation of the chemokine receptor CCR2-binding motif of monocyte chemoattractant protein-1 (MCP-1) and MCP-1 PAMs are evaluated preclinically as diagnostic tools for atherosclerosis. Monocyte-targeting is desirable as the influx of monocytes is a marker of early lesions, accumulation of monocytes is linked to atherosclerosis progression, and rupture-prone plaques have higher numbers of monocytes. MCP-1 PAMs bind to monocytes in vitro, and MCP-1 PAMs detect and discriminate between early- and late-stage atherosclerotic aortas. Moreover, MCP-1 PAMs are found to be eliminated via renal clearance and the mononuclear phagocyte system (MPS) without adverse side effects. Thus, MCP-1 PAMs are a promising new class of diagnostic agents capable of monitoring the progression of atherosclerosis.

In vivo biodistribution and clearance of peptide amphiphile micelles.

Peptide amphiphiles (PAs) are promising biomaterials for medical applications. To translate the use of PAs successfully from laboratories to clinics, in vivo studies regarding the safety of these nanomaterials are required. To examine the toxicity and clearance of PA biomaterials, we intravenously administered cy7-labeled, spherical PA micelles, control micelles without a peptide sequence, or PBS in a murine model and investigated biocompatibility, biodistribution, and clearance. Both peptide and non-peptide labeled micelles were approximately 8 nm in diameter, but of opposite surface charge. Neither micelle type caused aggregation or hemolysis of red blood cells. All micelles primarily accumulated in the bladder and were present in urine samples confirming elimination through renal clearance. Ex vivo imaging showed that micelles were also found in the liver suggesting some involvement of the reticuloendothelial system. However, no evidence of toxicity was found within the liver, spleen, kidney, bladder, intestines, lung, and heart. FROM THE CLINICAL EDITOR: Safety studies related to peptide amphiphile biomaterials are discussed in this paper, demonstrating that organotoxicity is unlikely with these materials, however, RES activation in the liver may be of consideration in further studies and needed for potential applications.

Active targeting of early and mid-stage atherosclerotic plaques using self-assembled peptide amphiphile micelles.

Inflammatory cell adhesion molecules expressed by endothelial cells on the luminal surface of atherosclerotic plaques, such as vascular cell adhesion molecule-1 (VCAM-1), provide a rational target for diagnostic and therapeutic delivery vehicles. Therefore, the potential of using spherical, self-assembled micelles synthesized from VCAM-1 targeted peptide amphiphile molecules was examined for the ability to specifically bind to both early and mid-stage atherosclerotic plaques. In vitro, cells incubated with VCAM-1 targeted and dye-labeled micelles show enhanced fluorescence signal as compared to cells incubated with a PEG micelle control. In vivo, VCAM-1 targeted and Cy7-labeled peptide amphiphile micelles were shown to specifically accumulate at atherosclerotic plaques in both early and mid-stage ApoE -/- mice through co-localization of Cy7 signal with anti-VCAM-1 antibody staining in fixed tissue. No specific accumulation was observed with a PEG micelle control. Histological analysis of excised tissue provided evidence for the in vivo biocompatibility of these micelle formulations as no tissue damage was observed. These results demonstrate that VCAM-1 targeted micelles have potential as a platform for targeted drug delivery to multiple stages of atherosclerotic plaque formation due to their established specificity and safety.