Executable biochemical space for specification and analysis of biochemical systems

• Published in: PloS one 2020-09, Vol.15 (9), p.e0238838-e0238838
• Main Authors: Troják, Matej, Šafránek, David, Mertová, Lukrécia, Brim, Luboš
• Format: Article
• Language: English
• Subjects: Algebra; Algorithms; Annotations; Automation; Biochemical Phenomena; Biochemistry; Biological research; Biology; Biology and Life Sciences; Chemical reactions; Cheminformatics; Chemistry; Computational biology; Computer and Information Sciences; Computer applications; Computer Simulation; Differential equations; Explosions; Genomes; High level languages; Humans; Integration; Laboratories; Language; Languages; Mathematical models; Methods; Models, Biological; Models, Theoretical; Ordinary differential equations; Phosphorylation; Physical Sciences; Programming Languages; Proteins - chemistry; Proteins - classification; Proteins - metabolism; Research and Analysis Methods; Robustness (mathematics); Semantics; Social Sciences; Software; Specifications; Syntax; Systems Biology
• ISSN: 1932-6203; 1932-6203
• DOI: 10.1371/journal.pone.0238838

Computational systems biology provides multiple formalisms for modelling of biochemical processes among which the rule-based approach is one of the most suitable. Its main advantage is a compact and precise mechanistic description of complex processes. However, state-of-the-art rule-based languages still suffer several shortcomings that limit their use in practice. In particular, the elementary (low-level) syntax and semantics of rule-based languages complicate model construction and maintenance for users outside computer science. On the other hand, mathematical models based on differential equations (ODEs) still make the most typical used modelling framework. In consequence, robust re-interpretation and integration of models are difficult, thus making the systems biology paradigm technically challenging. Though several high-level languages have been developed at the top of rule-based principles, none of them provides a satisfactory and complete solution for semi-automated description and annotation of heterogeneous biophysical processes integrated at the cellular level. We present the second generation of a rule-based language called Biochemical Space Language (BCSL) that combines the advantages of different approaches and thus makes an effort to overcome several problems of existing solutions. BCSL relies on the formal basis of the rule-based methodology while preserving user-friendly syntax of plain chemical equations. BCSL combines the following aspects: the level of abstraction that hides structural and quantitative details but yet gives a precise mechanistic view of systems dynamics; executable semantics allowing formal analysis and consistency checking at the level of the language; universality allowing the integration of different biochemical mechanisms; scalability and compactness of the specification; hierarchical specification and composability of chemical entities; and support for genome-scale annotation.