Towards a Scientific Model Management System

Computational models of biological systems aim at accurately simulating in vivo phenomena. They have become a very powerful tool enabling scientists to study complex behavior. A side effect of their success Unfortunately exists and is observed as an increasing difficulty in managing data, metadata and a myriad of programs and tools used and produced during a research task. In this work we aim at supporting scientists during a research endeavour by using Scientific Models as a main guiding element for describing, searching and running computational models, as well as managing the corresponding results. We assume a data-oriented perspective for scientific model representation materialized into a data model with which users describe scientific models and corresponding computational models, and a query language with which a scientist specifies simulation queries. The model is grounded in XML and tightly related to domain ontologies, which provide formal domain descriptions and uniform terminology. Scientists may search for scientific models and run simulations that automatically invoke the underlying programs on provided inputs. The results of a simulation may generate complex data that can be queried in the context of the scientific model. Higher-level models can be specified through views that export a unified representation of underlying scientific models.

Just-in-time Analytics Over Heterogeneous Data and Hardware

Manolis Karpathiotakis

Industry and academia are continuously becoming more data-driven and data-intensive, relying on the analysis of a wide variety of datasets to gain insights. At the same time, data variety increases continuously across multiple axes. First, data comes in multiple formats, such as the binary tabular data of a DBMS, raw textual files, and domain-specific formats. Second, different datasets follow different data models, such as the relational and the hierarchical one. Data location also varies: Some datasets reside in a central "data lake", whereas others lie in remote data sources. In addition, users execute widely different analysis tasks over all these data types. Finally, the process of gathering and integrating diverse datasets introduces several inconsistencies and redundancies in the data, such as duplicate entries for the same real-world concept. In summary, heterogeneity significantly affects the way data analysis is performed. In this thesis, we aim for data virtualization: Abstracting data out of its original form and manipulating it regardless of the way it is stored or structured, without a performance penalty. To achieve data virtualization, we design and implement systems that i) mask heterogeneity through the use of heterogeneity-aware, high-level building blocks and ii) offer fast responses through on-demand adaptation techniques. Regarding the high-level building blocks, we use a query language and algebra to handle multiple collection types, such as relations and hierarchies, express transformations between these collection types, as well as express complex data cleaning tasks over them. In addition, we design a location-aware compiler and optimizer that masks away the complexity of accessing multiple remote data sources. Regarding on-demand adaptation, we present a design to produce a new system per query. The design uses customization mechanisms that trigger runtime code generation to mimic the system most appropriate to answer a query fast: Query operators are thus created based on the query workload and the underlying data models; the data access layer is created based on the underlying data formats. In addition, we exploit emerging hardware by customizing the system implementation based on the available heterogeneous processors â CPUs and GPGPUs. We thus pair each workload with its ideal processor type. The end result is a just-in-time database system that is specific to the query, data, workload, and hardware instance. This thesis redesigns the data management stack to natively cater for data heterogeneity and exploit hardware heterogeneity. Instead of centralizing all relevant datasets, converting them to a single representation, and loading them in a monolithic, static, suboptimal system, our design embraces heterogeneity. Overall, our design decouples the type of performed analysis from the original data layout; users can perform their analysis across data stores, data models, and data formats, but at the same time experience the performance offered by a custom system that has been built on demand to serve their specific use case.

EPFL2017

Open Source Brain: A Collaborative Resource for Visualizing, Analyzing, Simulating, and Developing Standardized Models of Neurons and Circuits

Matteo Cantarelli, András Ecker, Michael Lee Hines, Sergio Solinas, Werner Alfons Hilda Van Geit

Computational models are powerful tools for exploring the properties of complex biological systems. In neuroscience, data-driven models of neural circuits that span multiple scales are increasingly being used to understand brain function in health and disease. But their adoption and reuse has been limited by the specialist knowledge required to evaluate and use them. To address this, we have developed Open Source Brain, a platform for sharing, viewing, analyzing, and simulating standardized models from different brain regions and species. Model structure and parameters can be automatically visualized and their dynamical properties explored through browser-based simulations. Infrastructure and tools for collaborative interaction, development, and testing are also provided. We demonstrate how existing components can be reused by constructing new models of inhibition-stabilized cortical networks that match recent experimental results. These features of Open Source Brain improve the accessibility, transparency, and repro-ducibility of models and facilitate their reuse by the wider community.

CELL PRESS2019

Just-in-time Analytics Over Heterogeneous Data and Hardware

Manolis Karpathiotakis

EPFL2017