Ahmed, Yasmine
(2022)
Towards the Automation of Expanding Dynamic Network Models with Knowledge from Literature.
Doctoral Dissertation, University of Pittsburgh.
(Unpublished)
Abstract
Creating computational models of world complex systems, including intracellular and intercellular bionetworks, geopolitical, economic, environmental, and agricultural world systems, is a time and labor-intensive task which is often limited by the knowledge and experience of modelers. This has naturally led to the emergence of the idea of automating the process of building new models or extending existing models, which could have a significant potential in enabling rapid, consistent, comprehensive and robust analysis of complex systems.
Inspired by this idea, we propose in this work different novel approaches for expanding models using the information extracted from literature using machine reading engines. Our proposed approaches combine machine reading with clustering, and graph theoretical analysis to create an automated framework for efficient model assembly. Furthermore, by automatically extending models with the information published in literature, our proposed methods allow for collecting the existing information in a consistent and comprehensive way. This, in turn, facilitates information reuse, data reproducibility, and replacing hundreds/thousands of manual experiments, thereby reducing the time needed for the advancement of knowledge.
We tested how well each method can reproduce manually built and curated models in different biological domains, when provided with varying amount of information in the baseline model and in the machine reading output. In particular, we have demonstrated the reliability of the proposed methods using three different selected models, namely, T cell differentiation, T cell large granular lymphocyte, and pancreatic cancer cell. Experimental results reveal considerable improvements of our approaches over other related methods. Moreover, using our automated model extension approach, we are able to efficiently find the best set of extensions to reproduce the manually extended models. Besides demonstrating automated reconstruction of a model that was previously built manually, our methods can assemble multiple models that satisfy desired system properties. As such, it replaces large number of tedious or even impractical manual experiments and guides alternative hypotheses and interventions in biological systems. Finally, we explored different model versions and system property testing results in order to develop a heuristic to modify model update rules.
Share
| Citation/Export: |
|
| Social Networking: |
|
Details
| Item Type: |
University of Pittsburgh ETD
|
| Status: |
Unpublished |
| Creators/Authors: |
|
| ETD Committee: |
|
| Date: |
10 June 2022 |
| Date Type: |
Publication |
| Defense Date: |
7 April 2022 |
| Approval Date: |
10 June 2022 |
| Submission Date: |
8 March 2022 |
| Access Restriction: |
No restriction; Release the ETD for access worldwide immediately. |
| Number of Pages: |
142 |
| Institution: |
University of Pittsburgh |
| Schools and Programs: |
Swanson School of Engineering > Electrical and Computer Engineering |
| Degree: |
PhD - Doctor of Philosophy |
| Thesis Type: |
Doctoral Dissertation |
| Refereed: |
Yes |
| Uncontrolled Keywords: |
Model extension, Model recommendation, Clustering, Model assembly, NLP, Collaboration graph, Markov clustering, Statistical model checking, Boolean modeling, Discrete modeling, Stochastic simulation, Model assembly, Automated modeling, network modeling. |
| Date Deposited: |
10 Jun 2022 19:25 |
| Last Modified: |
10 Jun 2022 19:25 |
| URI: |
http://d-scholarship.pitt.edu/id/eprint/42348 |
Metrics
Monthly Views for the past 3 years
Plum Analytics
Actions (login required)
 |
View Item |
![[feed]](/images/feed-icon-32x32.png){kind=icon}