Implementation of Allotrope Framework

Laboratories generate a significant amount of experimental data from a variety of sources – instruments, software, and human input. Since ages scientists/technicians in the lab have been spending a long time to maintain experiment data related papers and they can seem very productive with these papers. Laboratories need to be organized, maintained for multiple purposes like data retention guidelines as per regulatory compliances. Biologic linkage and tracing it back to the origination is of key concern to any scientist. Data is generated at every stage of an experiment e.g. from an ELN, while sequencing, from bio registers, during primary and secondary screenings, etc. This data must be immediately accessible for analysis as soon as the experiment is over. Every observation made is critical as it can help in a revolutionary innovation at some step of the other. There are many companies making software for input of data but for any given scientist the value of that data lies with its output. The proprietary data formats each instrument has made interchangeability of data and integration of different systems more difficult. A holistic option to connect all information, including meta-data, does not exist. Thus, scientists do not want to move away from paper for two main reasons: paper-driven procedures and a lack of well-integrated systems. While moving towards paperless laboratories what ideally needs to happen is to modify processes and ensure lesser dependency on papers. However, the modification of paper-driven procedures is not the only solution to achieve a paperless laboratory. Organizations also need to tackle the other significant root cause—the lack of an integrated system.

Today, the majority of laboratories are more or less automated—in the form of instrumentation and instrument data systems with Laboratory Information Management Systems (LIMS) being at the center. Typically, laboratories use many diverse types of software apart from LIMS. While LIMS is used to track sample lifecycle and related data management, the output of the analysis of the samples is performed through instruments interfacing. While trying to achieve “Paperless Flow” in laboratory, LIMS needs to be integrated with other enterprise software such as enterprise resource planning (ERP), electronic lab notebook (ELN), scientific data management systems (SDMS), chromatographic data systems (CDS), inventory management system, training management system, statistical package and so on. Though the intention is to have seamless interconnectivity between all these systems, in reality, many manual operations still prevail. Many times the workflow/data entry is done by a non-technical personnel or non-scientist. People working in the aggregator level of the umbrella don’t notice that query access to data and metadata generated by support processes is missing. Thus there is a disconnect in the data entry process and the data mining process. Most of the organizations are now trying to reduce the extent of manual operations and thereby move closer to the ideal paperless laboratory.

If you look at the instrumentation and analytical technologies in the market, they all come with embedded software. Increasingly, technologies in the pharmaceutical industry are being networked. U.S. Food and Drug Administration regulations, for example, require these instruments to be very tightly monitored and audited. So increasingly, the software part of instrumentation has become as important as the hardware part. When research becomes global, interconnectedness, collaboration, analytics at fingertips becomes a necessity. Thus regulatory compliance and business transformation objectives are the two drivers for the paperless laboratory.

This needs to be delivered by effective, efficient, data repositories, plus effective integration and data transfer between applications that constitute the paperless laboratory for an individual organization.

The issue of scientific data standardization and integration of laboratory elements has become a key concern for the players in the industry. There are various initiatives such as SiLA consortium (Standardization in Lab Automation), AnIML (Analytical Information Markup Language), Allotrope Foundation (ADF Framework) and Pistoia Alliance (HELM – single notation standard that can encode the structure of all biomolecules) to develop these common standards for the community.

Allotrope Foundation is an international consortium of pharmaceutical and biopharmaceutical companies with a common vision to develop innovative new standards and technology for handling data in R&D, with an initial focus on analytical chemistry. The effort of the Allotrope Foundation to create a common lab data format that is instrument and vendor “agnostic,” allowing for more efficient and compliant analytical and manufacturing control processes, aligns closely with FDA’s lab regulatory objectives, senior industry players involved are stressing upon. The Allotrope Framework is comprised of the Allotrope Data Format (ADF), taxonomies to provide a controlled vocabulary for metadata and a software toolkit. The ADF is a vendor agnostic format that stores data sets of unlimited size in a single file, organized as n-dimensional arrays in a data cube, and stores metadata describing the context of the equipment, process, materials, and results. The Framework enables cross-platform data transfer, data sharing, and vastly increases the ease of its use. This effort is fully funded by the members of Allotrope Foundation like Amgen, Bayer, Biogen, Pfizer, Baxter, etc. and is rapidly progressing towards achieving common goals to reduce wasted effort, improve data integrity while allowing realization of the value of analytical data.

The Framework is a toolkit that enables the consistent use of standards & metadata in software development currently comprised of three components and is designed to evolve as science and technology evolve, maintaining the access and interoperability with legacy data while lowering the barriers to innovation by removing the dependencies of legacy data formats.

1. 1. ADF:

The Allotrope Data Format (ADF) is a versatile data format capable of storing data sets of unlimited size in a single file in a vendor agnostic manner capable of handling any laboratory technique. This data can be easily stored, shared and used across operating systems. The ADF comprises a data cube for storing numerical data in n-dimensional arrays, a data description layer for storing contextual metadata in a Resource Description Framework (RDF) data model, and a data package that serves as a virtual file system to store ancillary files associated with an experiment. Class libraries are included in the Allotrope Framework to ensure the consistent adoption of the standards. The Foundation also provides a free ADF explorer – an application that can open any ADF file to view the data (data description, data cubes, data package) stored within. An ADF file details:

Why data was gathered (sample, study, purpose)

1. 1. • How that data was generated (instrument, method)
      • How data was processed (analysis method)
      • The shape of data (dimensions, measures, structure)

The ADF is intended to enable speedy real-time access to, and long-term stability of, archived analytical data. It has been designed to meet the performance requirements of advanced instrumentation, and be extensible by allowing new techniques and technologies to be incorporated while maintaining backward compatibility with previous versions.

1. 1. AFO:

The Allotrope Taxonomies and Ontology form the basis of a controlled vocabulary for the contextual metadata needed to describe and execute a test or measurement and later interpret the data. Drawing from thought leaders across member companies and the APN, the standard language for describing the equipment, processes, materials, and results are being developed to cover a broad range of techniques and instruments, driven by real use cases, in an extensible design.

1. 1. ADM:

Allotrope Data Models provide a mechanism to define data structures (schemas, templates) that describe how to use the ontologies for a given purpose in a standardized (i.e. reproducible, predictable, verifiable) way.

• Data Accessibility –The need for vendor-to-vendor technology integration is eliminated by creating an extensible data representation that facilitates easy access and sharing of the data output from any vendor’s software or laboratory equipment. This allows metadata, data in incompatible proprietary format, data in silos to be shared and accessed instantly.
• Data Integration – The Allotrope Framework’s standard format for data and metadata enables compatibility within laboratory infrastructure that will lower the effort and cost required to integrate applications and workflows. This will ensure higher system and process automation.
• Data Integrity – The Allotrope Framework addresses data integrity at the source by eliminating the need to convert between file formats or manually retype data, and prevent manual errors before they can occur.
• Regulatory Compliance – Interoperability within laboratory infrastructure allows linked Quality Control (QC) data and complete traceability of data over its full lifecycle. Adoption of the Allotrope Framework results in data that is easily read, searched and shared, effectively addressing data integrity and regulatory compliance issues.
• Scientific Reproducibility – The Framework enables a complete and accurate representation of the critical metadata needed to document experiments (methods, materials, conditions, results, algorithms) enabling reproducibility of the original work in just a few clicks.
• Improved Data Analytics – The Allotrope Framework dramatically improves the quality and completeness of metadata and reduces the time it takes to interconvert data between data sources. This allows for the successful implementation of big data and analytics strategy. Also, the data description layer of the ADF uses and RDF data model that provides the capability to build in business rules and other analytics on top of the standardized vocabularies.
• Reduced Costs – Ease of integration between laboratory equipment and software systems will serve to reduce IT expenditure by eliminating the need for customized solutions and software patches. Software and instrument interoperability will also lower effort and expense for support and maintenance. Also, the adoption of the Allotrope Framework allows more laboratory automation which will improve overall operational efficiency, leading to even more cost savings while laying down the foundation for innovations and new solutions in the data lifecycle. SME’s and Technology Partners

Member companies, collaborating with vendor partners, have begun to demonstrate how the framework enables cross-platform data transfer, facilitates finding, accessing and sharing data, and enables increased automation in laboratory data flow with a reduced need for error-prone manual input. Allotrope Foundation released the 1st phase of a framework for commercial use and is recognized with a 2017 Bio-IT World Best Practice Award.

As a part of the Allotrope Foundation, member companies are active in Allotrope working groups and teams, with a particular role, including teams defining technique specific taxonomies and data models, technical and ontology working groups, and defining governance and support processes. This collaboration amongst >100 diverse experts from the Pharmaceutical, Biopharmaceutical, Crop Sciences, Instrument and Software vendors in Analytical Sciences (Discovery, Development, and Manufacturing), Regulatory & Quality, Data Sciences, Information Technologies on an industry and cross-industry level, enables to oversee a wide range of technological trends and business needs.

Companies in the partner network like Abbott Informatics, Perkin Elmer, Agilent, Bio via, Labware, Metler Toledo, Terra science, Thermo Scientific, Waters, Persistent Systems, Shimadzu, etc. not only understand the holistic picture and the broader proposition of standardization they will be able to offer to their customers but will also play a role in the development of the standardized framework that can be practically implemented. Value of a particular data type or its application is significantly greater when shared than the same data in a silo.

Agilent is one of the members of the Allotrope Framework. Allotrope member companies have been engaged in the Allotrope Framework since 2012.

How Agilent contributes to the Allotrope Foundation organization

• Agilent’s chromatography software including Chemstation and MassHunter have data generated in their proprietary format.
• There is a rigorous need to standardize the data format for integration, while migration from Chemstation to MassHunter.
• The SIM ion from Agilent’s single quadrupole instrument has changed from a binary format (Chemstation) to an INI file format (Chemstation and MassHunter) and most recently an XML format (OpenLab 2). The terse format does not clearly indicate which number represents the SIM ion and which number is the dwell time. Furthermore, the unit of dwell time is unstated.
• Ultimately the ADF must be written and read in an environment that is commercially available to the supporters of the Allotrope Foundation.

To demonstrate this, a prototype software was developed that supports LC instruments and LC/MS single quadrupole Instruments on the ChemStation Edition of OpenLAB. The prototype consists of two components. The first component, the ChemStation2ADF converter, writes the ADF format with the method, raw data, results, instrument traces, and other metadata. Once the ADF is created, it is automatically uploaded to an OpenLAB Enterprise Content Management system (ECM) by the Scheduler. The second component, the ADF filter, reads the Data Description from the ADF and places the information into a relational database and which is immediately available to all users through the ECM search and retrieval mechanisms.

Future Activities

• Support other Mass Spectrometer types
• Include qualitative results
• Contribute to the ADF standard for MS
• Read ADFs produced by other vendors

To realize the benefits of the Allotrope framework, organizations will need to:

• Understand what the ADF, AFO, ADM are and how they are intended to be used

Current state/format of the data, including:

• Fitment of the Allotrope Framework into their project
• Plan
• Work with subject matter experts to define the desired shape of data (data description, data format, raw data)
• Work with Allotrope to understand how the Allotrope ontologies map to datasets
• Craft a staged project plan to move from their current state to the desired state

Plan for downstream uses of the data, including:

• Regulatory compliance
• Data archiving
• Define the processes and tools that will be used to convert data from its present state into Allotrope-compliant files
• Execute
• Train and support in-house resources
• Maintain and evolve the system as needs change
• Evolve
• With ever demanding needs of industry and compliances

As a member of the Allotrope Partner Network, Persistent Systems is uniquely positioned to assist your organization in developing and implementing an effective Allotrope Framework architecture and strategy. Digital Transformation and informatics are the central focus of our organization. Leveraging our scientific domain knowledge, technology expertise, industry experience, and extensive partner network, our professionals can help you across the product lifecycle services.

With the implementation of the Allotrope framework or the likes of data standardization frameworks, IT Systems in the laboratory will become more service oriented and plug and play allowing workflows to be software independent and vendor specific format agnostic, enabled by information, data and software standards. The use of data standards increases the interoperability of software tools, thus crafting the reality for a digital, intelligent and connected analytical laboratory. Apart from this, there will be better data synchronization, downstream process support will improve to enable rinse and repeat for follow-on processes. The Foundation aims to create an automated laboratory environment that drives better data analytics, one-click reports, scientific discovery and innovation, regulatory compliance, ultimately providing better medicines to patients faster.

The data management challenges that the Allotrope Framework is designed to address are certainly not exceptional to the Life Science Industry. Today, companies are exploring their digital potential and at the same time discovering facts they were unaware of earlier. Budget constraints, sustainability objectives, faster developments of products and services are important objectives for organizations. This is driving organizations to think about standardization and interoperability of technologies in the push to optimize innovation and maintain a competitive edge.