Mark Daniel Rintoul

Developing a methodology for an inverse quantitative structure-activity relationship using the signature molecular descriptor.

The concept of signature as a molecular descriptor is introduced and various topological indices used in quantitative structure-activity relationships (QSARs) are expressed as functions of the new descriptor. The effectiveness of signature versus commonly used descriptors in QSAR analysis is demonstrated by correlating the activities of 121 HIV-1 protease inhibitors. Our approach to the inverse-QSAR problem consists of first finding the optimum sets of descriptor values best matching a target activity and then generating a focused library of candidate structures from the solution set of descriptor values. Both steps are facilitated by the use of signature.

Constrained walks and self-avoiding walks: implications for protein structure determination

We prove that n-step walks and self-avoiding walks on the 2D honeycomb, 2D square, 3D diamond and 3D cubic lattices can be uniquely characterized (canonized) with no more than n Euclidian distances. We also demonstrate that these canonical distances can be obtained with O(n) physical measurements. Finally, while the protein-folding problem on lattices is known to be strongly NP-hard, we prove that lattice protein structures of size n matching O(n) canonical distance measurements can be determined in linear time.

Designing novel polymers with targeted properties using the signature molecular descriptor.

A method for solving the inverse quantitative structure-property relationship (QSPR) problem is presented which facilitates the design of novel polymers with targeted properties. Here, we demonstrate the efficacy of the approach using the targeted design of polymers exhibiting a desired glass transition temperature, heat capacity, and density. We present novel QSPRs based on the signature molecular descriptor capable of predicting glass transition temperature, heat capacity, density, molar volume, and cohesive energies of linear homopolymers with cross-validation squared correlation coefficients ranging between 0.81 and 0.95. Using these QSPRs, we show how the inverse problem can be solved to design poly(N-methyl hexamethylene sebacamide) despite the fact that the polymer was used not used in the training of this model.

A computational framework for ontologically storing and analyzing very large overhead image sets

We describe a computational approach to remote sensing image analysis that addresses many of the classic problems associated with storage, search, and query. This process starts by automatically annotating the fundamental objects in the image data set that will be used as a basis for an ontology, including both the objects (such as building, road, water, etc.) and their spatial and temporal relationships (is within 100 m of, is surrounded by, has changed in the past year, etc.). Data sets that can include multiple time slices of the same area are then processed using automated tools that reduce the images to the objects and relationships defined in an ontology based on the primitive objects, and this representation is stored in a geospatial-temporal semantic graph. Image searches are then defined in terms of the ontology (e.g. find a building greater than 103 m2 that borders a body of water), and the graph is searched for such relationships. This approach also enables the incorporation of non-image data that is related to the ontology. We demonstrate through an initial implementation of the entire system on large data sets (109 -- 1011 pixels) that this system is robust against variations in different image collection parameters, provides a way for analysts to query data sets in a more natural way, and can greatly reduce the memory footprint of the search.

Exploratory Trajectory Clustering with Distance Geometry

We present here an example of how a large, multi-dimensional unstructured data set, namely aircraft trajectories over the United States, can be analyzed using relatively straightforward unsupervised learning techniques. We begin by adding a rough structure to the trajectory data using the notion of distance geometry. This provides a very generic structure to the data that allows it to be indexed as an n-dimensional vector. We then do a clustering based on the HDBSCAN algorithm to both group flights with similar shapes and find outliers that have a relatively unique shape. Next, we expand the notion of geometric features to more specialized features and demonstrate the power of these features to solve specific problems. Finally, we highlight not just the power of the technique but also the speed and simplicity of the implementation by demonstrating them on very large data sets.

Trajectory analysis via a geometric feature space approach

This study aimed to organize a body of trajectories in order to identify, search for and classify both common and uncommon behaviors among objects such as aircraft and ships. Existing comparison functions such as the Frechet distance are computationally expensive and yield counterintuitive results in some cases. We propose an approach using feature vectors whose components represent succinctly the salient information in trajectories. These features incorporate basic information such as the total distance traveled and the distance between start/stop points as well as geometric features related to the properties of the convex hull, trajectory curvature and general distance geometry. Additionally, these features can generally be mapped easily to behaviors of interest to humans who are searching large databases. Most of these geometric features are invariant under rigid transformation. Furthermore, we demonstrate the use of different subsets of these features to identify trajectories similar to an exemplar, cluster a database of several hundred thousand trajectories and identify outliers.

Encoding and analyzing aerial imagery using geospatial semantic graphs

While collection capabilities have yielded an ever-increasing volume of aerial imagery, analytic techniques for identifying patterns in and extracting relevant information from this data have seriously lagged. The vast majority of imagery is never examined, due to a combination of the limited bandwidth of human analysts and limitations of existing analysis tools. In this report, we describe an alternative, novel approach to both encoding and analyzing aerial imagery, using the concept of a geospatial semantic graph. The advantages of our approach are twofold. First, intuitive templates can be easily specified in terms of the domain language in which an analyst converses. These templates can be used to automatically and efficiently search large graph databases, for specific patterns of interest. Second, unsupervised machine learning techniques can be applied to automatically identify patterns in the graph databases, exposing recurring motifs in imagery. We illustrate our approach using real-world data for Anne Arundel County, Maryland, and compare the performance of our approach to that of an expert human analyst.

Evaluating Parallel Relational Databases for Medical Data Analysis

Hospitals have always generated and consumed large amounts of data concerning patients, treatment and outcomes. As computers and networks have permeated the hospital environment it has become feasible to collect and organize all of this data. This raises naturally the question of how to deal with the resulting mountain of information. In this report we detail a proof-of-concept test using two commercially available parallel database systems to analyze a set of real, de-identified medical records. We examine database scalability as data sizes increase as well as responsiveness under load from multiple users.

Parallel Tempering Monte Carlo in LAMMPS

We present here the details of the implementation of the parallel tempering Monte Carlo technique into a LAMMPS, a heavily used massively parallel molecular dynamics code at Sandia. This technique allows for many replicas of a system to be run at different simulation temperatures. At various points in the simulation, configurations can be swapped between different temperature environments and then continued. This allows for large regions of energy space to be sampled very quickly, and allows for minimum energy configurations to emerge in very complex systems, such as large biomolecular systems. By including this algorithm into an existing code, we immediately gain all of the previous work that had been put into LAMMPS, and allow this technique to quickly be available to the entire Sandia and international LAMMPS community. Finally, we present an example of this code applied to folding a small protein.

The new biology and computational statistical physics

While it has historically been an exploratory, descriptive, and empirical science, in the past 100 years, biology has become more discovery- and mechanism-oriented. There are a number of ways in which this new paradigm is driving much of the current biological research toward statistical physics. This is happening at a molecular level due to the very large nature of biological molecules, such as proteins and nucleic acids. It is also occurring at the cellular level where random processes play an important role in cell function. There are even examples that describe the behavior of large numbers of individual organisms within one or more species. Finally, this trend has been accelerated with the advent of high-throughput experimental techniques that are driving biology towards information science. Analysis and discovery of the information gained from such experiments will rely heavily on techniques that have traditionally been applied in statistical physics. This paper will focus on examples of how statistical physics techniques are being applied and hope to be applied to biological problems, with an emphasis on high-performance computing. We will also speculate on what we feel are the necessary computing requirements to solve many of the outstanding problems in computational biology using the techniques that will be discussed.