Finding optimal candidate and composition of engineered biomaterials is a challenging task with significant experimental cost and effort. To design effective, de novo biomaterials for cardiac repair, this CAREER award supported by the Biomaterials program aims to investigate a new biomaterials design strategy with data-driven approaches. The first objective is to create a large number of functionally relevant proteins with a computer algorithm, which are subject to systematic screening and selection. The second objective is to assess the selected pool of candidate proteins in a 3D model tissue including healthy and infarct conditions, where the model tissue is 3D bio-printed with another computer algorithm to minimize experimental trials and effort. With data-driven approaches, the PI seeks to attain 1) optimal, newly designed sequences of therapeutic proteins to accelerate the proliferation of cardiomyocytes and 2) optimal combinations of biomaterials and 3D bio-printing parameters to fabricate the model tissue with minimal experimental cost. In contrast to conventional approaches, the proposed data-driven approaches are adaptive to altered experimental conditions to reach the optimal target of biomaterial properties with enhanced experimental efficiency. The outcome of this CAREER award can potentially provide a novel platform of engineering therapeutic proteins and fabricating a model tissue with optimal cost and effort. Outreach activities and educational curricula are proposed to engage a range of students from middle school to undergraduate students to promote the exposure of biomaterials research to those students in highly agricultural and petrochemical engineering-oriented industrial settings.
PART 2: TECHNICAL SUMMARY
This CAREER award aims to engineer an extracellular matrix (ECM) protein to accelerate the proliferation of cardiomyocytes. Once over 4 million functionally relevant chimeras are created by a computer algorithm, screening and selection of those chimeras are experimentally expensive. Thus, the new approach proposed in the CAREER award is to utilize publicly available data to build machine learning models to screen and select those chimeras, which will be built upon a Gaussian Process to derive a posterior from a prior and observations. To avoid the overfitting problem of machine learning with the relatively smaller number of publicly available data, an additional machine learning model will be built with 1) protein folding metrics and 2) intracellular protein-based assay. To assess the performance of designer ECM proteins, an in vitro model of infarct (border zone) will be fabricated with 3D bio-printing. To recapitulate complex features of the border zone with minimal numbers of trials of 3D bio-printing, Bayesian Optimization will be applied to balance the exploration of the overall search space and the exploitation of the local search space to effectively achieve the optimal 3D bio-printing and biomaterial parameters. To unambiguously identify the proliferation of cardiomyocytes in the model border zone, genetically labeled adult cardiomyocytes will be 3D bio-printed in the model border zone to assess the increase of cardiomyocyte proliferation. This CAREER award also seeks to support STEM education in the areas of biomaterials research. The outreach program targets specifically local secondary school and undergraduate students to attract them into future workforce in biomaterials areas with priorities to female and underrepresented groups in STEM education.
This project is jointly funded by the Biomaterials Program, Division of Materials Research, and the Established Program to Stimulate Competitive Research (EPSCoR).
This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.
