Home Page for Diane Uwacu | Parasol Laboratory


Picture Diane Uwacu
PhD Student

Parasol Laboratory url: http://parasollab.web.illinois.edu/~duwacu/
Department of Computer Science email:
University of Illinois at Urbana-Champaign office: 3307 Siebel Center
Urbana, IL 61801, USA


CV Google Scholar

I am a visiting PhD student at the University of Illinois from Texas A&M University, and I work on motion planning algorithms with Dr. Nancy Amato.
I started my research journey in the Parasol lab through the CRA-WP DREU program in the summer of 2014. For 10 weeks, I worked on reinforcement learning methods to improve local planning.
My current research interests are in motion planning applied in computational biology and robotics. Specifically, I work on topology-guided methods to predict accessibility.

Research

Skeleton-guided motion planning


We use prior knowledge of the workspace to efficiently sample relevant regions in the high dimensional space of the movable object

Protein-ligand Binding Site Accessibility Analysis


We use Rapidly-exploring Random Graphs coupled with Mean Curve workspace skeletons to find valid paths for ligand motion into the protein binding site

Adaptive planning


We applied reinforcement learning to reward motion planning tools based on their performance in heterogeneous environments. This work was tested on robotics and computational biology applications.

Publications

Topology-Guided Roadmap Construction with Dynamic Region Sampling, Read Sandström, Diane Uwacu, Jory Denny, Nancy M. Amato, IEEE Robotics and Automation Letters (RA-L), vol. 5, no. 4,, pp. 6161-6168, virtual, Oct 2020. DOI: 10.1109/LRA.2020.3010487
Keywords: Sampling-Based Motion Planning, Workspace Topology
Links : [Published]

BibTex

@ARTICLE{9144398,
author={R. {Sandström} and D. {Uwacu} and J. {Denny} and N. M. {Amato}},
journal={IEEE Robotics and Automation Letters},
title={Topology-Guided Roadmap Construction With Dynamic Region Sampling},
year={2020},
volume={5},
number={4},
pages={6161-6168},}


Abstract

Many types of planning problems require discovery of multiple pathways through the environment, such as multi-robot coordination or protein ligand binding. The Probabilistic Roadmap (PRM) algorithm is a powerful tool for this case, but often cannot efficiently connect the roadmap in the presence of narrow passages. In this letter, we present a guidance mechanism that encourages the rapid construction of well-connected roadmaps with PRM methods. We leverage a topological skeleton of the workspace to track the algorithm's progress in both covering and connecting distinct neighborhoods, and employ this information to focus computation on the uncovered and unconnected regions. We demonstrate how this guidance improves PRM's efficiency in building a roadmap that can answer multiple queries in both robotics and protein ligand binding applications.


Using Guided Motion Planning to Study Binding Site Accessibility, Diane Uwacu, Abigail Ren, Shawna Thomas, Nancy M. Amato, Proceedings of the 11th ACM International Conference on Bioinformatics, Computational Biology and Health Informatics, Issue: 109, (Virtual) New York, USA, Sep 2020. DOI: 10.1145/3388440.3414707
Keywords: Computational Biology, Ligand Binding, Motion Planning
Links : [Published] [Manuscript]

BibTex

@inbook{10.1145/3388440.3414707,
author = {Uwacu, Diane and Ren, Abigail and Thomas, Shawna and Amato, Nancy M.},
title = {Using Guided Motion Planning to Study Binding Site Accessibility},
year = {2020},
isbn = {9781450379649},
publisher = {Association for Computing Machinery},
address = {New York, NY, USA},
url = {https://doi.org/10.1145/3388440.3414707},
abstract = {Computational methods are commonly used to predict protein-ligand interactions. These methods typically search for regions with favorable energy that geometrically fit the ligand, and then rank them as potential binding sites. While this general strategy can provide good predictions in some cases, it does not do well when the binding site is not accessible to the ligand. In addition, recent research has shown that in some cases protein access tunnels play a major role in the activity and stability of the protein's binding interactions. Hence, to fully understand the binding behavior of such proteins, it is imperative to identify and study their access tunnels. In this work, we present a motion planning algorithm that scores protein binding site accessibility for a particular ligand. This method can be used to screen ligand candidates for a protein by eliminating those that cannot access the binding site. This method was tested on two case studies to analyze effects of modifying a protein's access tunnels to increase activity and/or stability as well as study how a ligand inhibitor blocks access to the protein binding site.},
booktitle = {Proceedings of the 11th ACM International Conference on Bioinformatics, Computational Biology and Health Informatics},
articleno = {109},
numpages = {10}
}




Abstract

Computational methods are commonly used to predict protein-ligand interactions. These methods typically search for regions with favorable energy that geometrically fit the ligand, and then rank them as potential binding sites. While this general strategy can provide good predictions in some cases, it does not do well when the binding site is not accessible to the ligand. In addition, recent research has shown that in some cases protein access tunnels play a major role in the activity and stability of the protein's binding interactions. Hence, to fully understand the binding behavior of such proteins, it is imperative to identify and study their access tunnels. In this work, we present a motion planning algorithm that scores protein binding site accessibility for a particular ligand. This method can be used to screen ligand candidates for a protein by eliminating those that cannot access the binding site. This method was tested on two case studies to analyze effects of modifying a protein's access tunnels to increase activity and/or stability as well as study how a ligand inhibitor blocks access to the protein binding site.


Using Motion Planning to Evaluate Protein Binding Site Accessibility, Diane Uwacu, Everett Yang, Shawna Thomas, Nancy M. Amato, Technical Report, TR18-001, Parasol Laboratory, Department of Computer Science, Texas A&M University, College Station TX 77848, USA, Jul 2018.
Keywords: Ligand Binding, Motion Planning, Workspace Topology
Links : [Manuscript]

BibTex

N/A


Abstract

Despite many efforts and considerable breakthroughs in ligand binding prediction, the best predictors still produce many false positives and a reliable, fully automated prediction framework has yet to be developed. Binding site accessibility is an important feature ignored by methods that classify binding based solely on the energetic or geometric properties of the final bound protein-ligand complex. To evaluate this necessity, we transform the ligand accessibility problem into a robot motion planning problem where the ligand is modeled as a flexible agent whose task is to travel from outside the protein to its binding site. We use Rapidly-exploring Random Graphs coupled with Mean Curve workspace skeletons to quickly and thoroughly explore a protein environment in order to produce valid paths for ligand motion. Path weights reflect the influences of intermolecular forces on the given ligand. Low weight paths are extracted and analyzed for characteristics of accessibility. In this paper, we show that our algorithm provides a mechanism to evaluate binding site accessibility for a ligand.


The Impact of Approximate Methods on Local Learning in Motion Planning, Diane Uwacu, Chinwe Ekenna, Shawna Thomas, Nancy Amato, 1st International Workshop on Robot Learning and Planning (RLP 2016), in conjunction with RSS 2016, pp. 38-44, Ann Arbor, MI, Jun 2016.
Keywords: Machine Learning, Sampling-Based Motion Planning
Links : [ArXiv] [Manuscript]

BibTex

@inproceedings{uwacu-wrlp-2016
, author = "Diane Uwacu and Chinwe Ekenna and Shawna Thomas and Nancy M. Amato"
, title = "The Impact of Approximate Methods on Local Learning in Motion Planning"
, booktitle = "1st International Workshop on Robot Learning and Planning (RLP 2016), in conjunction with RSS 2016"
, month = "June"
, year = "2016"
, note = "http://chitsazlab.org/robotics/rlp2016/RLP_2016_paper_11.html"
}


Abstract

Machine learning methods have been applied to many motion planning algorithms including probabilistic roadmap methods (PRM). There are many variants of these methods and choosing the best one every time is hard and depends on local properties of the environment. A successful learning approach has been developed to offset this issue. This learning approach was applied to PRMs to help decide intelligently what method to utilize in dynamically created local regions of the environment or task space. It used exact neighbor finding approaches and removed the need to partition environments to get improved results. In this work we make further advances by introducing approximate neighbor finder methods. It has been established that approximate neighbor finding methods are faster than exact methods, still work well in connecting nodes to edges in PRMs, and that connection is robust to noise. We study what happens when noise is introduced into learning by using approximate methods instead of already studied exact methods. We show that the impact of noise on learning depends on how much learning needs to take place given the topology of the environment. Our results demonstrate a correlation between heterogeneity and the need for learning over a local region.


Improved Roadmap Connection via Local Learning for Sampling Based Planners, Chinwe Ekenna, Diane Uwacu, Shawna Thomas, Nancy Amato, Proc. IEEE/RSJ Int. Conf. Intel. Rob. Syst. (IROS), pp. 3227-3234, Hamburg, Germany, Oct 2015. DOI: 10.1109/IROS.2015.7353825
Keywords: Machine Learning, Motion Planning, Sampling-Based Motion Planning
Links : [Published]

BibTex

@INPROCEEDINGS{7353825, author={C. {Ekenna} and D. {Uwacu} and S. {Thomas} and N. M. {Amato}}, booktitle={2015 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)}, title={Improved roadmap connection via local learning for sampling based planners}, year={2015}, volume={}, number={}, pages={3227-3234},}


Abstract

Probabilistic Roadmap Methods (PRMs) solve the motion planing problem by constructing a roadmap (or graph) that models the motion space when feasible local motions exist. PRMs and variants contain several phases during roadmap generation i.e., sampling, connection, and query. Some work has been done to apply machine learning to the connection phase to decide which variant to employ, but it uses a global learning approach that is inefficient in heterogeneous situations. We present an algorithm that instead uses local learning: it only considers the performance history in the vicinity of the current connection attempt and uses this information to select good candidates for connection. It thus removes any need to explicitly partition the environment which is burdensome and typically difficult to do. Our results show that our method learns and adapts in heterogeneous environments, including a KUKA youBot with a fixed and mobile base. It finds solution paths faster for single and multi-query scenarios and builds roadmaps with better coverage and connectivity given a fixed amount of time in a wide variety of input problems. In all cases, our method outperforms the previous adaptive connection method and is comparable or better than the best individual method.


Studying Learning Techniques in Different Phases of PRM Construction, Chinwe Ekenna, Diane Uwacu, Shawna Thomas, Nancy Amato, In Machine Learning in Planning and Control of Robot Motion Workshop (IROS-MLPC), Hamburg, Germany, Oct 2015. DOI: 10.1109/IROS.2015.7353825
Keywords: Machine Learning, Sampling-Based Motion Planning
Links : [Published]

BibTex

no bib for now.


Abstract

Probabilistic Roadmap Methods (PRMs) solve the motion planning problem in two phases by sampling free configurations and connecting them together to build a map that is used to find a valid path. Existing algorithms are highly sensitive to the topology of the problem, and their efficiency depends on applying them to a compatible problem. Reinforcement learning has been applied to motion planning and rewards the action performed by planners during either sampling or connection, but not both. Previous work computed a global reward and action scheme, which saw a setback when heterogeneous environments were concerned. Local learning (connection) was recently introduced to offset this weakness identified during global learning, and there was some improvement in planner performance. These different learning schemes (global and local) have shown strengths and weaknesses individually. In this paper, we investigate local learning for sampling. We study what type of learning to apply when, and how the two phases of PRM roadmap construction interact, which has not been investigated before. We show the performance using each scheme on a KUKAyouBot, an 8 degree of freedom robot, and analyze what happens when they are all combined during roadmap construction