Publishing reproducible dynamic kinetic models.

Publishing repeatable and reproducible computational models is a crucial aspect of the scientific method in computational biology and one that is often forgotten in the rush to publish. The pressures of academic life and the lack of any reward system at institutions, granting agencies and journals means that publishing reproducible science is often either non-existent or, at best, presented in the form of an incomplete description. In the article, we will focus on repeatability and reproducibility in the systems biology field where a great many published models cannot be reproduced and in many cases even repeated. This review describes the current landscape of software tooling, model repositories, model standards and best practices for publishing repeatable and reproducible kinetic models. The review also discusses possible future remedies including working more closely with journals to help reviewers and editors ensure that published kinetic models are at minimum, repeatable. Contact: hsauro@uw.edu.

High-grade serous ovarian cancer cell lines exhibit heterogeneous responses to growth factor stimulation.

BACKGROUND: The factors driving the onset and progression of ovarian cancer are not well understood. Recent reports have identified cell lines that are representative of the genomic pattern of high-grade serous ovarian cancer (HGSOC), in which greater than 90 % of tumors have a mutation in TP53. However, many of these representative cell lines have not been widely used so it is unclear if these cell lines capture the variability that is characteristic of the disease. METHODS: We investigated six TP53-mutant HGSOC cell lines (Caov3, Caov4, OV90, OVCA432, OVCAR3, and OVCAR4) for migration, MMP2 expression, proliferation, and VEGF secretion, behaviors that play critical roles in tumor progression. In addition to comparing baseline variation between the cell lines, we determined how these behaviors changed in response to four growth factors implicated in ovarian cancer progression: HB-EGF, NRG1ß, IGF1, and HGF. RESULTS: Baseline levels of each behavior varied across the cell lines and this variation was comparable to that seen in tumors. All four growth factors impacted cell proliferation or VEGF secretion, and HB-EGF, NRG1ß, and HGF impacted wound closure or MMP2 expression in at least two cell lines. Growth factor-induced responses demonstrated substantial heterogeneity, with cell lines sensitive to all four growth factors, a subset of the growth factors, or none of the growth factors, depending on the response of interest. Principal component analysis demonstrated that the data clustered together based on cell line rather than growth factor identity, suggesting that response is dependent on intrinsic qualities of the tumor cell rather than the growth factor. CONCLUSIONS: Significant variation was seen among the cell lines, consistent with the heterogeneity of HGSOC.

A Practical Guide to Reproducible Modeling for Biochemical Networks.

While scientific disciplines revere reproducibility, many studies - experimental and computational alike - fall short of this ideal and cannot be reproduced or even repeated when the model is shared. For computational modeling of biochemical networks, there is a dearth of formal training and resources available describing how to practically implement reproducible methods, despite a wealth of existing tools and formats which could be used to support reproducibility. This chapter points the reader to useful software tools and standardized formats that support reproducible modeling of biochemical networks and provides suggestions on how to implement reproducible methods in practice. Many of the suggestions encourage readers to use best practices from the software development community in order to automate, test, and version control their model components. A Jupyter Notebook demonstrating several of the key steps in building a reproducible biochemical network model is included to supplement the recommendations in the text.

Standards, dissemination, and best practices in systems biology.

The reproducibility of scientific research is crucial to the success of the scientific method. Here, we review the current best practices when publishing mechanistic models in systems biology. We recommend, where possible, to use software engineering strategies such as testing, verification, validation, documentation, versioning, iterative development, and continuous integration. In addition, adhering to the Findable, Accessible, Interoperable, and Reusable modeling principles allows other scientists to collaborate and build off of each other's work. Existing standards such as Systems Biology Markup Language, CellML, or Simulation Experiment Description Markup Language can greatly improve the likelihood that a published model is reproducible, especially if such models are deposited in well-established model repositories. Where models are published in executable programming languages, the source code and their data should be published as open-source in public code repositories together with any documentation and testing code. For complex models, we recommend container-based solutions where any software dependencies and the run-time context can be easily replicated.

Best Practices for Making Reproducible Biochemical Models.

Like many scientific disciplines, dynamical biochemical modeling is hindered by irreproducible results. This limits the utility of biochemical models by making them difficult to understand, trust, or reuse. We comprehensively list the best practices that biochemical modelers should follow to build reproducible biochemical model artifacts-all data, model descriptions, and custom software used by the model-that can be understood and reused. The best practices provide advice for all steps of a typical biochemical modeling workflow in which a modeler collects data; constructs, trains, simulates, and validates the model; uses the predictions of a model to advance knowledge; and publicly shares the model artifacts. The best practices emphasize the benefits obtained by using standard tools and formats and provides guidance to modelers who do not or cannot use standards in some stages of their modeling workflow. Adoption of these best practices will enhance the ability of researchers to reproduce, understand, and reuse biochemical models.

Targeting Ligands Deliver Model Drug Cargo into the Central Nervous System along Autonomic Neurons.

While biologic drugs such as proteins, peptides, or nucleic acids have shown promise in the treatment of neurodegenerative diseases, the blood-brain barrier (BBB) severely limits drug delivery to the central nervous system (CNS) after systemic administration. Consequently, drug delivery challenges preclude biological drug candidates from the clinical armamentarium. In order to target drug delivery and uptake into to the CNS, we used an in vivo phage display screen to identify peptides able to target drug-uptake by the vast array of neurons of the autonomic nervous system (ANS). Using next-generation sequencing, we identified 21 candidate targeted ANS-to-CNS uptake ligands (TACL) that enriched bacteriophage accumulation and delivered protein-cargo into the CNS after intraperitoneal (IP) administration. The series of TACL peptides were synthesized and tested for their ability to deliver a model enzyme (NeutrAvidin-horseradish peroxidase fusion) to the brain and spinal cord. Three TACL-peptides facilitated significant active enzyme delivery into the CNS, with limited accumulation in off-target organs. Peptide structure and serum stability is increased when internal cysteine residues are cyclized by perfluoroarylation with decafluorobiphenyl, which increased delivery to the CNS further. TACL-peptide was demonstrated to localize in parasympathetic ganglia neurons in addition to neuronal structures in the hindbrain and spinal cord. By targeting uptake into ANS neurons, we demonstrate the potential for TACL-peptides to bypass the blood-brain barrier and deliver a model drug into the brain and spinal cord.

The Extracellular Matrix of Ovarian Cortical Inclusion Cysts Modulates Invasion of Fallopian Tube Epithelial Cells.

A growing body of research supports the idea that the fallopian tube epithelium (FTE) is the precursor for most high-grade serous ovarian canacers (HGSOC) but that the ovary plays a critical role in tumor metastasis. Cortical inclusion cysts (CICs) in the ovarian cortex have been hypothesized to create a niche environment that plays a role in HGSOC progression. Through histological analysis of pathology samples from human ovaries, we determined that collagen I and III were elevated near CICs and that the collagen fibers in this dense region were oriented parallel to the cyst boundary. Using this information from human samples as design parameters, we engineered an in vitro model that recreates the size, shape, and extracellular matrix (ECM) properties of CICs. We found that FTE cells within our model underwent robust invasion that was responsive to stimulation with follicular fluid, while ovarian surface epithelial (OSE) cells, the native cells of the ovary, were not invasive. We provide experimental evidence to support a role of the extracellular matrix in modulating FTE cell invasion, as decreased collagen I concentration or the addition of collagen III to the matrix surrounding FTE cells increased FTE cell invasion. Taken together, we show that an in vitro model of CICs informed by the analysis of human tissue can act as an important tool for understanding FTE cell interactions with their environment.