Completing and Adapting Models of Biological Processes

Abstract

Abstract. We present a learning-based method for model completion and adaptation, which is based on the combination of two approaches: 1) R2D2C, a technique for mechanically transforming system requirements via provably equivalent models to running code, and 2) automata learning-bafied model extrapolation. The intended impact of this new combination is to make model completion and adaptation accessible to experts of the field, like biologists or engineers. The principle is briefly illustrated by generating models of biological procedures concerning gene activities in the production of proteins, although the main application is going to concern autonomic systems for space exploration. MotivationA formal approach to Requirements-Based Programming, provisionally named R2D2C ("Requirements to Design to Code"), was developed at NASA [1] as a general-purpose method to mechanically transform system requirements into a provably equivalent model. This is a central need for ultra-high dependability systems like those developed at NASA for space exploration. The R2D2C approach provides mathematically tractable round-trip engineering for system development, rigorously based on formal modelling and formal reasoning techniques. In this paper we complement this method with a learning-based method for model completion and adaptation in order to make model completion and adaptation accessible to experts of the field, like biologists or engineers.Before discussing the technical background and the biological application, we briefly sketch the standard areas of application. Application AreasThe work described below is motivated by the need for requirements-based programming for ultra-high dependability systems which are remote, embedded, and increasingly autonomic.

Bio-jETI: a service integration, design, and provisioning platform for orchestrated bioinformatics processes

“…In [ 47 ] we presented a learning-based method for model completion and adaptation, which is based on the combination of two approaches: 1) R2D2C, a NASA proprietary technique for mechanically transforming system requirements via provably equivalent models to running code, and 2) automata learning-based model extrapolation as provided in the LearnLib, which is part of the FMICS-jETI platform. The intended impact of this new combination is to make model completion and adaptation accessible to experts of the field, like biologists or engineers.…”

CONNECT Challenges: Towards Emergent Connectors for Eternal Networked Systems

“…To respect the difference between input and output events, inference of Mealy machines has been developed by the University of Dortmund [48]: the underlying principles of inference algorithms remain the same. More recent work concerns relational model construction [49], synthesis of design models from scenarios, with human interaction [50,71], and further optimizations of regular inference techniques for assume-guarantee reasoning [51].…”

From the How to the What

“…-complex workflow for retrieving orthologous promoters [14], -the analysis with statistical methods of large data sets from the LC/MS analysis for protein identification and discovery [9], -the service-oriented redesign of GeneFisher, a popular and successful tool for PCR primer design [10], and -the derivation of models for cellular processes by means of active learning techniques [13].…”