This Small Business Innovation Research Program (SBIR) Phase I project will develop algorithms and prototype software to track and quantify forces experienced by single-biomolecules in living cells. The approach is built on recent advances in time series analysis to: (1) fit 2-D/3-D stochastic differential equations that accurately characterize the underlying particle kinetics (methods for checking models against experimental data will be provided); and, (2) form reliable tracks from crowded and noisy image sequences containing many molecules by extending state-of-the-art algorithms from target tracking applications. Such analysis tools do not exist in current single particle tracking software. The tools will be utilized to extract new kinetic information on protein motion in the primary cilium of mouse cells due to the system's relevance in biophysics, cell signaling, and cancer research. The algorithms and software will provide more accurate estimates of kinetic parameters with the unique addition of goodness-of-fit hypothesis testing metrics.
The broader impact/commercial potential of this project will be the development of new software tools capable of producing powerful insights into cellular and synthetic biological systems by enabling the extraction of novel information characterizing bimolecular motion in live cells. Such insights will positively influence numerous research areas ranging from enhanced drug delivery to improved yields in synthetic biology applications. The commercial software will enable researchers with various backgrounds to: (1) produce reliable tracks from large image sequences (with metrics describing the quality of the candidate tracks); (2) extract the underlying kinetic information; and, (3) test the assumptions behind standard biophysical models. The algorithms and software developed by this effort will serve as the basis for a unique commercial product offering that will meet unmet demand in the biological imaging market.
Problem Background: Advances in microscopy have equipped scientists with tools capable of measuring the motion of individual molecules (i.e., proteins, DNA, RNA, etc.) inside of living cells with high spatial and temporal resolution. Prior to these advances, many scientists were constrained to resolution limits imposed by the diffraction barrier of light (i.e., a spatial resolution of roughly 200 nanometers). This resolution limit has hindered scientists’ ability to monitor the motion of individual molecules in their native cellular environment (single molecules in the cell are typically much smaller than 200 nanometers). Various super-resolution microscopy techniques have recently overcome the diffraction limit. Numerous labs now are routinely producing high volumes of data characterizing molecular motion since quantitative information about the motion of individual molecules in the cell allows scientists to address many open problems in biophysics and molecular biology. However, statistical techniques capable of efficiently extracting the wealth of molecular information detectable by modern microscopes are under-developed. Reliably extracting scientific information from super-resolution measurements poses exciting challenges at the interface of several disciplines. Intellectual Merit: During the six month Phase I SBIR grant, we developed new time series and tracking algorithms for reliably extracting quantitative information characterizing the motion of molecules in living cells from measurements produced by single-molecule imaging experiments. New algorithms for estimating parameters of 2D and 3D stochastic models were developed. Statistical techniques for checking assumptions implicit in the fitted/assumed observations were also generated (these techniques addressed technical complications facing real-world laboratory measurements). Computational routines for enhancing and extending "track association" were also produced. These algorithms leveraged new ideas coming from Statistics, Applied Mathematics, and Mathematical Finance to address open issues associated with analyzing complex experimental super-resolution data. The new algorithms were tested on both simulation [1] and experimental [2-3] data. In addition to the aforementioned publications, the ideas underlying the algorithms have been publicly presented at several academic conferences and seminars. During Phase I, the primary focus was on applying our new algorithms and computational tools to biological systems of interest to practicing scientists. The aim was to demonstrate the utility of our envisioned software product when it is applied to assist in analyzing experimental single-molecule data. In terms of specific applications studied, we developed new methods for characterizing protein transport in the primary cilium in a mouse cell line. The primary cilium is an organelle found in nearly all mammalian cells. This organelle is believed to be important in organizing several signaling cascades related to normal fundamental development and various diseases including cancer. However, the mechanisms by which proteins are transported into and out of this organelle are not currently understood. The single-molecule analysis techniques explored in this Phase I work demonstrated promise in giving new insights about the nature of diffusion and directed transport experienced by proteins in the primary cilium of live mammalian cells [3]. In addition, we have begun modifying and applying the computational tools to quantify force generation and sensing by the mitotic spindle during cell division. We have proposed to continue this work in our Phase II SBIR effort since a molecular level understanding of cell division is a long-standing , important, and open problem. Broader Impacts: The algorithms have been implemented in a prototype software package which can wrap around popular image processing tools currently in the standard workflow of numerous researchers. Trial software licenses have been issued to multiple researchers at two universities (Colorado State U. and U. North Carolina). Agreements between other leading institutions are currently being negotiated. The computational tools developed can help in characterizing molecular motion in a variety of applications, ranging from genetically engineering yeast to understanding fundamental molecular processes like cell division. Therefore, providing these computational tools to researchers aiming to expand the frontier of single-molecule techniques shows enormous potential in accelerating the scientific discovery process. It is our hope that feedback from these early adopters will catalyze and enhance the development of a successful commercial software package that can assist numerous academic and industry researchers aiming to use optical microscopic measurements to aid in the understanding of molecular motion within live cells. References: [1] Calderon, Phys. Rev. E (2013). [2] Calderon, Thompson, Casolari, Paffenroth, Moerner, J. Phys. Chem. B (2013). [3] Calderon, Weiss, Moerner, (submitted; http://arxiv.org/abs/1312.6742).
