EvoNEST, a modular application for comprehensive species-based sampling and data management in comparative biology

Introduction

The field of ecology and comparative biology is increasingly data-intensive, with researchers collecting and processing information on organisms, their traits, and their environments (Stucky et al., 2018; Perez-Riverol et al., 2019; Schoch et al., 2020; Pekár et al., 2021; Herberstein et al., 2022; Gloor et al., 2023; Bánki et al., 2023; Daru, 2024). This rapid increase in data availability has substantially advanced our understanding of biodiversity, species interactions, and evolutionary processes. However, it has also presented substantial challenges in data management, particularly for research groups working with diverse species and the samples and traits derived from them (Jones et al., 2006; Michener & Jones, 2012).

Many laboratories and institutions in biodiversity research face common hurdles in data management: researchers often use a variety of tools and formats to record data, leading to inconsistencies and difficulties in data integration (Michener, 2015; Poisot et al., 2019); critical data is frequently scattered across different devices, file formats, and users, hampering collaboration and data retrieval (Wilkinson et al., 2016). As research projects grow in complexity, integrating interdisciplinary approaches that generate diverse data types, maintaining complete records of sample origins, treatments, and derived data becomes increasingly challenging (Roche et al., 2015). Without centralized and standardized data management systems, there is a greater risk of losing information due to human error or technical issues (Hughes-Lartey et al., 2021). Researchers often spend significant time on data entry, retrieval, and formatting, reducing time available for analysis and interpretation (Boettiger et al., 2015).

These challenges are particularly pressing in the current research landscape, where there are increasing demands for data digitization, complete documentation for regulatory compliance (e.g., collection licenses, animal welfare legislation), and comprehensive data sharing for publication (Whitlock, 2011). This challenge has been discussed for more than fifteen years, highlighting how ecology could benefit from a systematic, better use of standardisation methods (Kelling et al., 2009; Reichman, Jones & Schildhauer, 2011; Hampton et al., 2013; Powers & Hampton, 2019; König et al., 2019). As an approach for the digital accessibility of data published in public repositories, in 2016, the Findability, Accessibility, Interoperability, Reusability (FAIR) principles were proposed (Wilkinson et al., 2016). These principles are now a fundamental part of open science practices and aim to provide guidelines to improve the findability, accessibility, interoperability, and reuse of digital assets (Wilkinson et al., 2016). While literature exist on how to apply FAIR data principles to large cloud databases (Mons et al., 2017; Jacobsen et al., 2020; van Reisen et al., 2020; The Galaxy Community et al., 2022), there is a lack of discussion about how local data could be prepared in each ecology and comparative biology laboratory according to those principles (Borer et al., 2009; White et al., 2013), while more general works highlight how open data often does not have the necessary data quality (Nikiforova, 2020). Often a variety of files, especially calculation sheets (such as Excel spreadsheets) structured in individual, non-standardized ways, are used by different lab members. These sheets are then adapted and restructured several times depending on the current need, i.e., an analysis, a data exportation, or an upload to an online database.

To address these challenges, we have developed Evolutionary, ecological and biological Nexus of Experiments, Samples and Traits (EvoNEST), a modular full-stack application designed to streamline the collection, management, exploratory analysis, and preparation for publication of data in ecological and evolutionary research. EvoNEST provides a centralized system for storing, processing, and analysing organism (specimen, sample, and subsample) data, measurements, and raw data files. EvoNEST aims to improve researchers’ efficiency, while ensuring that the data natively respect the FAIR principles.

Although initially developed for a comparative study on spiders and spider silk properties, EvoNEST has been designed with flexibility in mind, allowing for adaptation to a wide range of organism-based research contexts. The platform can accommodate various types of organisms, from plants to animals, and can be customised to handle different types of samples, traits, and experimental data. To facilitate adoption across different research contexts, we provide a backbone version of EvoNEST that can be easily installed and customised while maintaining all core functionalities.

Compared to already existing platforms sharing similar missions (Gries, Gilbert & Franz, 2014; Popović et al., 2020; Bogan et al., 2023), EvoNEST provides a local solution which can be installed within the digital boundaries of the laboratory. EvoNEST focuses on providing the tools to optimise not only the collection, but also the management and analysis of the organisms. These advantages allow new and existing laboratories to standardise their local data management. The data architecture does not have a rigid schema, allowing facilities to customise their own platform and data structure, share their data, while keeping ownership of their data. A similar approach has been successfully implemented before by OpenBioMaps, a data management platform dedicated to spatial biodiversity data (Bán et al., 2022).

In this article, we present the architecture and key features of EvoNEST, demonstrating how it addresses common challenges faced by comparative biologists collecting samples and data across many species. We outline how storing and managing data through EvoNEST allows for a more transparent data handling, and how EvoNEST implements FAIR data principles. Finally, we compare EvoNEST with existing solutions, and we present, as a case study, an implementation based on a comparative study on spiders.

Methods and software overview

EvoNEST is a database creator, manager and visualiser full-stack application based on Node.js (https://nodejs.org/). Like any Node.js application, it can be run locally on a machine, or deployed on a server. In both cases, it is accessible and navigable via a browser. EvoNEST’s frontend is built using the Next.js framework. The backend is written using Next.js (https://nextjs.org/) and MongoDB (https://www.mongodb.com/). Data visualisation is granted by a variety of data visualisation tools, including customisable plots and tables, based on both self-written components and external libraries, cited in our repository. A schematic overview about the software design is found in Fig. 1. The entire architecture is containerized using Docker, allowing rapid installation and customisation (Merkel, 2014).

Case study: comparative study of spider traits

To demonstrate EvoNEST’s capabilities, we show a fully implemented NEST focusing on comparative spider research. This implementation showcases both the backbone version’s core functionalities and customised features developed for spider and silk research. A simplified entity-relationship diagram of this implementation is shown in Fig. 2.

Building upon the foundations of the backbone version, we developed specialized features for spider and spider silk research. We implemented automatic taxonomic data retrieval and validation through integration with the World Spider Catalog (WSC) online database. This not only allows for the automatic addition of higher taxonomic entities (e.g., family assignments), and the automatic correction of misspelled species names, but also the automatic update of taxonomic changes to sample species. The subsample management system was customised to handle silk samples, with specialised categorisation by silk sample type (e.g., dragline, capture thread). Usually, multiple silk samples can be derived from a single animal, and, further, it is possible that silk samples are further divides into silk subsamples for different analyses, all of which is accounted for by EvoNEST’s nested database structure.

For the import and processing of experimental data, we extended the backbone’s experiment collection capabilities to process output files from a KLA T-150 universal testing machine that was used to measure the mechanical properties of spider silk samples. Custom modules automatically extract and calculate key mechanical properties such as peak strength, extensibility, and toughness, and store these as traits. Traits acquired through direct measurements (such as the body mass of spiders) can also be added manually to the trait list. In addition, diverse files generated through observations and experiments, such as documents and images, can be stored and connected with sample and trait entries. The dashboard and some visualisation tools were customised to highlight important traits and trait correlations measured in the lab and, to enable project monitoring and exploratory data analyses.

This implementation is available as a demo at https://evonest.zoologie.uni-greifswald.de/. In a version of this application, currently used in our lab (Fig. 3), twelve active users contribute to the main laboratory NEST, managing at the time of writing over 4,100 samples, 700 experiments, and 7,800 traits, demonstrating the system’s scalability for specialized research contexts.

Comparison with existing solutions

Several systems have been developed to address challenges in biodiversity and laboratory data management. Known solutions are, for example: Electronic Lab Notebooks (ELN) (Kanza et al., 2017), LIMS (Laboratory Information Management Systems) applications for biodiversity research, self-deployable solutions like OpenBioMaps (Bán et al., 2022), and museum-focused platforms such as Symbiota (Gries, Gilbert & Franz, 2014). However, EvoNEST addresses a specific gap in this ecosystem that existing systems do not effectively fill. A comparison between EvoNEST and existing solutions is available in Table 2.

Traditional LIMS platforms focus primarily on sample tracking and workflow management in clinical or manufacturing environments, with licensing costs ranging from $3,000 to $250,000+ annually and implementation complexity requiring dedicated IT resources. While powerful for enterprise environments, these systems require extensive customisation to handle organism-based research workflows and taxonomic hierarchies that are fundamental to biological collections. Electronic Laboratory Notebooks (ELNs) provide flexible documentation capabilities and offer many benefits over physical lab notebooks, including networked digital environments, collaboration support, and contribution to Good Research Practice through tracking and documentation of research processes (Vandendorpe et al., 2024). However, as ELNs are primarily designed as digital replacements for traditional laboratory notebooks rather than extensive data management systems for complex specimen relationships and taxonomic hierarchies, they face inherent limitations when applied to systematic organism-based research. Researchers often maintain separate systems for specimen tracking and experimental data analysis, creating parallel data streams with attendant risks of inconsistency and duplication (Vandendorpe et al., 2024). Unlike these platforms that focus primarily on sample tracking and workflow management, EvoNEST was designed from the ground up for organism-based research with inherent support for taxonomic hierarchies, parent-child relationships between specimens, and the complex web of derived samples and measurements.

More biodiversity-specific solutions are rich in features and require less training, but lack flexibility and are often more suited for collections rather than active experimental research environments: OpenBioMaps and Symbiota serve, respectively, the spatial biodiversity monitoring community, and natural history museums and herbaria. While each software excels in specimen cataloguing, taxonomic databases, and collaborative digitization, they lack integrated living culture and specimen maintenance tracking, document data processing, and the trait analysis capabilities required for organism-based research.

Design consideration and development roadmap

The development of EvoNEST involved several technical decisions that shaped its current implementation. The decision with the greatest impact on the application development regarded the data structure. We chose MongoDB over relational databases to accommodate heterogeneous data types with variable attributes across different organisms and research contexts. This document-oriented approach allows researchers to add new fields or data types without restructuring the database, supporting the exploratory nature of ecological and evolutionary research. Although MongoDB’s document-oriented approach introduces some computational overhead compared to optimised relational queries, we determined this cost to be minimal relative to the flexibility gains. This flexibility has proven valuable for diverse research applications, as it allows developers to quickly create new front-end components dedicated to the management of their specific organism. For each modular component of EvoNEST, we provide template files that can be used as reference. This is described in our developer guide, available at https://daniele-liprandi.github.io/EvoNEST-backbone/developer-guide.

We plan to further enhance EvoNEST through several upcoming developments, including:

• Standardised data exchange formats and direct integration with major biodiversity repositories. This will include semi-automated data submission to platforms like Global Biodiversity Information Facility (GBIF), through the use of custom data transformers and, when available, repository APIs.

• Specialised front-end modules for different research domains (e.g., plant phenotyping, microbial cultures, longitudinal ecological studies).

• Implementation of locally-installed large language models (LLMs) to provide contextual assistance and data exploration capabilities. This will allow non-expert users such as students and early-career researchers to query complex datasets using natural language, generate preliminary analyses, and receive guidance on data collection protocols without requiring advanced database or programming skills.

To continue EvoNEST’s development through community involvement, we have adopted an open-source strategy. Collaborating on the development of EvoNEST is possible through the EvoNEST Backbone’s repository, at the URL https://github.com/daniele-liprandi/EvoNEST-backbone. The user guide, the developer guide, and a static version of the API Docs are all available at https://daniele-liprandi.github.io/EvoNEST-backbone/. We are establishing a community-driven expansion model where researchers can contribute specialised modules or modifications tailored to specific research contexts. This approach has proven successful in other scientific software ecosystems, such as R/Bioconductor (Huber et al., 2015) and Galaxy (The Galaxy Community et al., 2022).

Conclusions

We have shown here that EvoNEST has broad applicability across various domains of ecological and evolutionary research. It is particularly suited for comparative biology studies that collect data across organisms and species, as well as long-term ecological monitoring and integrative biodiversity research. The platform’s versatility allows researchers to use it for a wide range of organisms, including microorganisms, plants and animals, in both field and laboratory settings. While basic knowledge of server management is beneficial for network deployment, the backbone version’s Docker-based installation process makes EvoNEST accessible to users with minimal technical expertise. Once installed, the system can be operated and customised by users without a programming background.

One of the main features of EvoNEST is its schema-less database design. We chose to follow this strategy to satisfy two needs of organism-based research: the existence of a multitude of data structures, and the growing necessity of organising and preserving the data. Both these needs exist in active research projects and when preparing data for publication, as the scientific community and the funding agencies require more and more transparency over the years. While this design choice sacrifices some computational efficiency when compared to more structured databases, these costs are offset by the ease of use that is gained. This allows EvoNEST to be easily implemented in every laboratory, without requiring the large amount of ad-hoc customisation needed by other existing alternatives, as ELNs and LIMS.

The goal of our application is to democratise access to biodiversity data, preserve data privacy and ownership, and increase the efficiency of laboratories dealing with different types of organismal data. To achieve these objectives, we invite interested scientific communities to engage with EvoNEST. Scientists and developers can contribute to the development of the app by providing feedback and feature requests, or by helping directly in the development of the app or its plugins and modules. All these actions can be taken through our GitHub repository. A collaborative approach to the development of EvoNEST will ensure a continuous evolution of its features, allowing it to meet the diverse and changing needs of the research community.

As data management challenges in biodiversity, ecology and evolution continue to grow, tools like EvoNEST give researchers the means to effectively organise, analyse, and share their data. Furthermore, having a unified platform for data collection, data storage and organism maintenance helps laboratories establish collaborative experiments, without sacrificing the ability to quickly run their analysis. By aiding the digitisation of phenotypic data, a digitisation that is still lagging behind the one of genetic data, EvoNEST may contribute to the advancement of our understanding of biodiversity and evolutionary processes.