Kirk E. Jordan
Kirk E. Jordan has a total of 25 co-authors. Here are the top co-authors by number of shared publications:
- Thomas J. Peters
University of Connecticut
5 shared publications - Ji Li
Department of Mathematics University of Connecticut 196 Auditorium Road Unit 3009 06269 Storrs CT
3 shared publications - D. Marsh
Pratt and Whitney
2 shared publications - Alexander Russell
University of Connecticut
1 shared publication - Chang Sik Kim
The Hartree Centre, STFC Daresbury Laboratory, Warrington, WA4 4AD, UK
1 shared publication - Connor W. Coley
Massachusetts Institute of Technology
1 shared publication
Latest Publications
- Open Access
- 11 December 2023
- Published by American Chemical Society (ACS) in ChemRxiv
A Physics-Inspired Approach to the Understanding of Molecular Representations and Models
The story of machine learning in general, and its application to molecular design in particular, has been a tale of evolving representations of data. Understanding the implications of the use of a particular representation -- including the existence of so-called `activity cliffs' for cheminformatics models -- is the key to their successful use for molecular discovery. In this work we present a physics-inspired methodology which exploits analogies between model response surfaces and energy landscapes to richly describe the relationship between the representation and the model. From these similarities, a metric emerges which is analogous to the commonly used frustration metric from the chemical physics community. This new property shows state-of-the-art prediction of model error, whilst belonging to a novel class of roughness measure that extends beyond the known data allowing the trivial identification of activity cliffs even in the absence of related training or evaluation data.
- 11 December 2023
- Published by American Chemical Society (ACS) in ChemRxiv
Open AccessA Physics-Inspired Approach to the Understanding of Molecular Representations and Models
The story of machine learning in general, and its application to molecular design in particular, has been a tale of evolving representations of data. Understanding the implications of the use of a particular representation -- including the existence of so-called `activity cliffs' for cheminformatics models -- is the key to their successful use for molecular discovery. In this work we present a physics-inspired methodology which exploits analogies between model response surfaces and energy landscapes to richly describe the relationship between the representation and the model. From these similarities, a metric emerges which is analogous to the commonly used frustration metric from the chemical physics community. This new property shows state-of-the-art prediction of model error, whilst belonging to a novel class of roughness measure that extends beyond the known data allowing the trivial identification of activity cliffs even in the absence of related training or evaluation data.
- Open Access
Knot Morphing Algorithm for Quantum `Fragile Topology'
A knot theoretic algorithm is proposed to model `fragile topology' of quantum physics.
Open AccessKnot Morphing Algorithm for Quantum `Fragile Topology'
A knot theoretic algorithm is proposed to model `fragile topology' of quantum physics.
- Open Access
K-mer clustering algorithm using a MapReduce framework: application to the parallelization of the Inchworm module of Trinity
Background: De novo transcriptome assembly is an important technique for understanding gene expression in non-model organisms. Many de novo assemblers using the de Bruijn graph of a set of the RNA sequences rely on in-memory representation of this graph. However, current methods analyse the complete set of read-derived k-mer sequence at once, resulting in the need for computer hardware with large shared memory. Results: We introduce a novel approach that clusters k-mers as the first step. The clusters correspond to small sets of gene products, which can be processed quickly to give candidate transcripts. We implement the clustering step using the MapReduce approach for parallelising the analysis of large datasets, which enables the use of compute clusters. The computational task is distributed across the compute system, and no specialised hardware is required. Using this approach, we have re-implemented the Inchworm module from the widely used Trinity pipeline, and tested the method in the context of the full Trinity pipeline. Validation tests on a range of real datasets show large reductions in the runtime and per-node memory requirements, when making use of a compute cluster. Conclusions: Our study shows that MapReduce-based clustering has great potential for distributing challenging sequencing problems, without loss of accuracy. Although we have focussed on the Trinity package, we propose that such clustering is a useful initial step for other assembly pipelines.
Open AccessK-mer clustering algorithm using a MapReduce framework: application to the parallelization of the Inchworm module of Trinity
Background: De novo transcriptome assembly is an important technique for understanding gene expression in non-model organisms. Many de novo assemblers using the de Bruijn graph of a set of the RNA sequences rely on in-memory representation of this graph. However, current methods analyse the complete set of read-derived k-mer sequence at once, resulting in the need for computer hardware with large shared memory. Results: We introduce a novel approach that clusters k-mers as the first step. The clusters correspond to small sets of gene products, which can be processed quickly to give candidate transcripts. We implement the clustering step using the MapReduce approach for parallelising the analysis of large datasets, which enables the use of compute clusters. The computational task is distributed across the compute system, and no specialised hardware is required. Using this approach, we have re-implemented the Inchworm module from the widely used Trinity pipeline, and tested the method in the context of the full Trinity pipeline. Validation tests on a range of real datasets show large reductions in the runtime and per-node memory requirements, when making use of a compute cluster. Conclusions: Our study shows that MapReduce-based clustering has great potential for distributing challenging sequencing problems, without loss of accuracy. Although we have focussed on the Trinity package, we propose that such clustering is a useful initial step for other assembly pipelines.
- 18 April 2015
- Published by Springer Nature
- p. 3-23
Algebraic Multigrid on a Dragonfly Network: First Experiences on a Cray XC30
- 18 April 2015
- Published by Springer Nature
- p. 3-23
Algebraic Multigrid on a Dragonfly Network: First Experiences on a Cray XC30
- Open Access
- 1 January 2014
- Published by European Alliance for Innovation n.o.
Exploring HPC-based Scientific Software as a Service using CometCloud
The use of in-silico simulations in experimental science can greatly increase laboratory efficiency and provide additional insights into interactions not easily described by traditional methods. Such simulations require significant amounts of computational resources, accessible only via supercomputers of large-scale high-performance clusters. Due to the complexity of the computational experiments, as well as the usage of the underlying resources, experimental scientists heavily rely on computational scientists with HPC expertise to perform these simulations. This additional bottleneck prevents the widespread adoption of real time in-silico simulation as a driver for laboratory experimentation. In this paper, we aim to overcome this bottleneck by presenting the architecture of an end-to-end framework to enable HPC Software as a Service. This framework is designed to make it easy for scientific applications to run on top of dynamically federated HPC resources. The framework enables HPC resource sharing while maximizing throughput and utilization. We focus specifically on a use case where an experimental scientist uses a mobile portal to control dissipative particle dynamics experiments that are executed on a remote supercomputer (IBM Blue Gene/Q).
- 1 January 2014
- Published by European Alliance for Innovation n.o.
Open AccessExploring HPC-based Scientific Software as a Service using CometCloud
The use of in-silico simulations in experimental science can greatly increase laboratory efficiency and provide additional insights into interactions not easily described by traditional methods. Such simulations require significant amounts of computational resources, accessible only via supercomputers of large-scale high-performance clusters. Due to the complexity of the computational experiments, as well as the usage of the underlying resources, experimental scientists heavily rely on computational scientists with HPC expertise to perform these simulations. This additional bottleneck prevents the widespread adoption of real time in-silico simulation as a driver for laboratory experimentation. In this paper, we aim to overcome this bottleneck by presenting the architecture of an end-to-end framework to enable HPC Software as a Service. This framework is designed to make it easy for scientific applications to run on top of dynamically federated HPC resources. The framework enables HPC resource sharing while maximizing throughput and utilization. We focus specifically on a use case where an experimental scientist uses a mobile portal to control dissipative particle dynamics experiments that are executed on a remote supercomputer (IBM Blue Gene/Q).
- Open Access
Computational Topology for Approximations of Knots
The preservation of ambient isotopic equivalence under piecewise linear (PL) approximation for smooth knots are prominent in molecular modeling and simulation. Sufficient conditions are given regarding: (1) Hausdorff distance, and (2) a sum of total curvature and derivative. High degree Bezier curves are often used as smooth representations, where computational efficiency is a practical concern. Subdivision can produce PL approximations for a given B\'ezier curve, fulfilling the above two conditions. The primary contributions are: (i) a priori bounds on the number of subdivision iterations sufficient to achieve a PL approximation that is ambient isotopic to the original Bezier curve, and (ii) improved iteration bounds over those previously established.
Open AccessComputational Topology for Approximations of Knots
The preservation of ambient isotopic equivalence under piecewise linear (PL) approximation for smooth knots are prominent in molecular modeling and simulation. Sufficient conditions are given regarding: (1) Hausdorff distance, and (2) a sum of total curvature and derivative. High degree Bezier curves are often used as smooth representations, where computational efficiency is a practical concern. Subdivision can produce PL approximations for a given B\'ezier curve, fulfilling the above two conditions. The primary contributions are: (i) a priori bounds on the number of subdivision iterations sufficient to achieve a PL approximation that is ambient isotopic to the original Bezier curve, and (ii) improved iteration bounds over those previously established.
- Open Access
- 29 July 2013
- Published by Universitat Politecnica de Valencia in Applied General Topology
- Vol. 13 (2) , 115-134
Computational Topology Counterexamples with 3D Visualization of Bézier Curves
Computational Topology Counterexamples with 3D Visualization of Bézier Curves
- 29 July 2013
- Published by Universitat Politecnica de Valencia in Applied General Topology
- Vol. 13 (2) , 115-134
Open AccessComputational Topology Counterexamples with 3D Visualization of Bézier Curves
Computational Topology Counterexamples with 3D Visualization of Bézier Curves
- Open Access
Knot Visualization Experiments for Verifiable Molecular Movies
Classical topological concepts are applied to understand high performance computing simulations of molecules writhing in three dimensional space. These simulations produce peta-bytes of floating point data, to describe 3 dimensional changes in molecular structure. A zero-th order analysis is achieved by viewing a computer animation synchronized with these changes. The performance demands for animation of this voluminous data can become problematic, but techniques from low-dimensional topology are helpful. The 3D molecule is reduced to a lower dimensional model of a 1-manifold, which undergoes a piecewise linear approximation for animation. An example is presented here to show how a 1-manifold and its PL approximation could come to have different embeddings as molecular writhing proceeds. This should serve as a cautionary warning to animators to respect established sufficient conditions for topological preservation so that the movies generated will faithfully reflect the topology of the underlying model. To obtain this result, techniques from low-dimensional topology were joined with experimental mathematics and numerical analyses.
Open AccessKnot Visualization Experiments for Verifiable Molecular Movies
Classical topological concepts are applied to understand high performance computing simulations of molecules writhing in three dimensional space. These simulations produce peta-bytes of floating point data, to describe 3 dimensional changes in molecular structure. A zero-th order analysis is achieved by viewing a computer animation synchronized with these changes. The performance demands for animation of this voluminous data can become problematic, but techniques from low-dimensional topology are helpful. The 3D molecule is reduced to a lower dimensional model of a 1-manifold, which undergoes a piecewise linear approximation for animation. An example is presented here to show how a 1-manifold and its PL approximation could come to have different embeddings as molecular writhing proceeds. This should serve as a cautionary warning to animators to respect established sufficient conditions for topological preservation so that the movies generated will faithfully reflect the topology of the underlying model. To obtain this result, techniques from low-dimensional topology were joined with experimental mathematics and numerical analyses.
- Open Access
Computational Topology Counterexamples with 3D Visualization of Bezier Curves
For applications in computing, Bezier curves are pervasive and are defined by a piecewise linear curve L which is embedded in R^3 and yields a smooth polynomial curve C embedded in R^3. It is of interest to understand when L and C have the same embeddings. One class of counterexamples is shown for L being unknotted, while C is knotted. Another class of counterexamples is created where L is equilateral and simple, while C is self-intersecting. These counterexamples were discovered using curve visualizing software and numerical algorithms that produce general procedures to create more examples.
Open AccessComputational Topology Counterexamples with 3D Visualization of Bezier Curves
For applications in computing, Bezier curves are pervasive and are defined by a piecewise linear curve L which is embedded in R^3 and yields a smooth polynomial curve C embedded in R^3. It is of interest to understand when L and C have the same embeddings. One class of counterexamples is shown for L being unknotted, while C is knotted. Another class of counterexamples is created where L is equilateral and simple, while C is self-intersecting. These counterexamples were discovered using curve visualizing software and numerical algorithms that produce general procedures to create more examples.