ACD Labs Logo


March 6-10, 2016
Georgia World Congress Center, Atlanta, GA, USA

Related Materials

Related presentations, posters, and scientific talks from this event have been posted here for your reference. Please click the associated link to download.

A New Approach to the Automated Identification of Metabolites in Multi-Vendor Datasets, R. Lee, V. Lashin, A. Paramonov, A. Sakharov

ACD/Labs Highlights its Mixture Analysis Software Capabilities at PITTCON 2016, ACD/Labs
ACD/Labs Highlights its Mixture Analysis Software Capabilities at PITTCON 2016 - March 2016

Meet ACD/Labs Staff

Booth # 4058

Graham McGibbon, Manager, Scientific Solutions

Michael Boruta, Industrial Solutions & Optical Spectroscopy Product Manager

Richard Lee, Solutions Manager, Metabolite Identification & Characterization

Sanji Bhal, Director of Marketing and Communications

Natalie Koay, Account Manager

Brent Pautler, Technical Specialist

Conference Networking Session

Sunday, March 6th, 1:30 PM

Measure and/or Predict? Perspectives on Prediction Approaches with and Versus Experiments for R & D Compound Analyses
Graham A. McGibbon and Sanji K. Bhal

Abstract: View Abstract

Effectively preparing, isolating, characterizing and understanding compounds is an often costly, time-consuming but vital effort of many chemical industries, including pharmaceutical, food, beverage, cosmetic, and lubricant companies and their ingredient suppliers. Approaches to ease the burden by predicting various chemical properties, ADME, toxicological, spectroscopic and chromatographic characteristics are thus inevitably of interest. Advances in laboratory hardware, software, and informatics technologies are all changing the way in which scientists work and share data and information with internal and external partners. For instance, improving the overall productivity of drug discovery, development and acceptance requires a combination of strategies, but even moderate improvements can substantially increase returns [1].

Rapid classification, identification, characterization, or elucidation of newly synthesized compounds and related or unexpected metabolites, impurities, contaminants, degradants and adulterants is a methodically intense undertaking. Software-aided interpretation tools can help speed up analysis and in silico screening may be employed to assess the potential impact various impurities or metabolites of a potential drug may have in order to prioritize further testing when toxicology data is limited or lacking [2].

The goal of this session is to share experiences in this ever-changing landscape, of lessons learned in using predictive software tools, noting areas of existing and emerging reliability and where pitfalls remain. Best practices and new perspectives on predictive technologies that come out of this dialog can be leveraged to aid scientists and their colleagues in a multi-disciplinary environments.

We invite attendees from all industries considering and especially actively using predictions in their workflows to participate and offer their perspectives and current challenges.

[1] Eric David, Tony Tramontin and Rodney Zemmel, The Road to Positive Returns, Invention Reinvented: McKinsey perspectives on pharmaceutical R&D, McKinsey & Company (2010)
[2] Gary E. Martin, Chapter 5: A Systematic Approach to Impurity Identification, Analysis of Drug Impurities, April 30, 2007.

Oral Presentations

Wednesday, March 9th, 8:50 AM

Informatics for Externalization
Graham A. McGibbon
Abstract Number: 1660-2
Session 1660: LIMS-No One Size Fits All (Room B405)

Abstract: View Abstract

Challenges in data sharing and collaboration are already apparent throughout most chemically involved R&D industries and being exacerbated by a trend towards outsourcing core and non-core scientific activities. It has been estimated that pharmaceutical and biotech R&D spending outside of company boundaries is now >40% [1] and Contract Research Organizations (CROs) in China, e.g. Wuxi and ChemPartner, are among the largest employers of synthetic chemists. As externalization and research virtualization evolve and business critical tasks of library synthesis, process chemistry, metabolism, toxicology, and intermediates manufacturing are included, the demand for information accessibility intensifies. Typical data sharing, by email on demand or via document management systems like Sharepoint™ between each Sponsor Company and their CROs isnt yet optimal. Formats are diverse and many, even spreadsheets and PDF files, may be less apt for future leveraging in a scientifically meaningful way.

This presentation will highlight a new laboratory informatics externalization model based on analytical data sharing as a use case. A standardized way to collaborate systematically and efficiently via accessing share data is suggested. A software platform ensures data is managed in a way that enables data mining for the purpose of identifying raw materials, impurities, metabolites and other chemical ingredients. Collaborative workspaces enable creating analytical knowledge packages that can have 'live' analytical data, metadata, and chemistry information independent of instrument source for seamless sharing. The value of fingertip access to this 'live' information that can be searched, shared, re-processed, re-purposed, and re-analyzed will be discussed. Platform knowledge content can be easily accessed via web or mobile client interfaces and is amenable to integration and cloud-based deployments.

[1] M.E. Elliot, The De-Evolution of Informatics, Scientific Computing, October 2012.

Poster Session

Wednesday, March 9th, Morning Session, 400 Aisle

A New Approach to the Automated Identification of Metabolites in Multi-Vendor Datasets
Richard Lee
Abstract Number: 1770-P1
Session 1770: Pharmaceutical-MS, LC/MS and Others

Abstract: View Abstract

For scientists involved in the study of drug metabolism the challenge of accurately and rapidly identifying metabolites is met using a variety of techniques and software packages. Whilst most instruments vendors offer software to help identify metabolites they are of little use in a multi-vendor laboratory, and they lack the flexibility and customization required. Most users have to then resort to purchasing additional software and using Microsoft Excel to get a complete workflow.

Here we present a new approach for the automated identification of metabolites, which allows data from nearly all the mass spectrometry vendors to be processed, reviewed and databased. The automated file capture and processing capabilities makes it suitable for high-throughput environment whilst the biotransformation scientist can check the results and make any changes such as modifying an assignment. The batch processing allows multiple time point samples to be processed and the calculation of pharmacokinetic parameters such as area under the cure (AUC). To help reduce the number of false positives a structure based prediction approach is used. Confirmation of the site of biotransformation is checked using the available MSMS data. The metabolite fragment mass shifts, relative to the parent MSMS spectrum, help localize the site of biotransformation. In the cases where there is not sufficient evidence to support a single site of biotransformation, then the metabolite structure can be represented using the Markush notation. Unexpected metabolites are also be identified using a combination of mass defect filtering, control sample comparison and component profiling. All the metabolites and metadata can then be stored in a database for future use. This allows for a greater degree of collaboration between the discovery and development departments which can save a huge amount of time and effort. Answering the question Have I seen this metabolite before? then becomes very easy.