Automation Of Structure Analysis In Pharmaceutical R&D
Gary A. McClusky, Brian Tobias, Robert Short, Dana DeJohn, Leslie McMacken, Carmen Faraianu, and Bryan Judge
Analytical Research Section, Chemistry Department
Parke-Davis Pharmaceutical Research
Division of Warner-Lambert Company
Ann Arbor, Michigan 48105
Note: This paper was originally presented at The 1996 International Symposia on Laboratory Automation and Robotics (ISLAR96). The support of Zymark in presenting this work in NetSci is gratefully acknowledged.
http://www.netsci.org/Science/LIMS/feature02.html
Abstract
The analytical chemistry laboratory in pharmaceutical research continues to be challenged to perform structural characterization rapidly and accurately. This is fueled by improvements in organic synthesis, such as automated synthesis and combinatorial chemistry, and the availability of better structural tools and methods. Advances in instrumentation, robotics and computer technology have enabled automation of the associated tasks, such as sample preparation, sample introduction, data acquisition and processing, and data storage and retrieval. The integration of these components with laboratory information systems results in improved overall efficiency of laboratory operation. This, in turn, allows the laboratory analyst to direct significant time to challenging structural characterization projects and away from standard analysis.
Laboratory robotics have been used to automate sample preparation for NMR and MS analysis. Bar-coded sample vials are delivered to the robot and are prepared using information obtained from the laboratory information management system (LIMS). Modules associated with this system include bar-code reader, solvent dispensing station, vortex mixer, pipetter, and turbidity check and filtration stations. The dissolved sample solution is dispensed into an NMR tube or MS vial, which is then capped and placed into an autosampler rack on the respective spectrometer.
Sample analysis is performed in a batch mode on automated NMR and mass spectrometers using instrument worklists downloaded from the LIMS. Data acquisition and processing occur without analyst intervention. Data file transfer and loading completed results into the LIMS are triggered by the analyst. Data files are archived on a file server equipped with an optical jukebox and indexed in a relational database. Data files are retrieved through querying the database and transferring files back to a spectrometer. Information system features include sample login front-end for customer use, email delivery of results, intranet access to sample and instrument status, and results through Netscape. The careful integration of analytical instrumentation, robotics and information systems results in an efficiently operating analytical laboratory that can meet the ever increasing demands of a research environment.
Introduction
Advancements in analytical instrumentation, computer technology and robotics have enabled the analytical chemistry laboratory to efficiently meet the challenge of supporting drug discovery programs in the pharmaceutical industry. The Analytical Research laboratory at Parke-Davis provides a full range of services and research in analytical chemistry. These include such diverse disciplines as separations science, three dimensional structural studies of biological targets and ligands, bioanalytical characterization studies, and structural chemistry. Automation has had a tremendous impact in providing structure characterization support of our synthetic chemistry effort. The primary structure methods used to support synthetic chemistry are NMR spectroscopy, mass spectrometry, infrared spectroscopy, and microanalysis. The greatest success and impact in automation has been achieved in the areas of NMR spectroscopy and mass spectrometry.
Our efforts to automate analytical services for NMR spectroscopy and mass spectrometry are described in this paper. Automated methods for sample preparation and sample introduction for both NMR spectroscopy and mass spectrometry were developed. Integration of these automated methods with our laboratory information management system (LIMS) produced numerous benefits and further increased efficiency of operation. Automated methods to store spectral data files were developed also.
Experimental
Sample Preparation: A Sagian ORCA robot is used to prepare samples for analysis by NMR spectroscopy and mass spectrometry. A three-meter rail configuration is used and works on one-side of the rail so as to be contained on a lab bench. Several custom modules were incorporated to provide the needed functionality. These modules include sample feeder racks, bar-code reader, vortex station, capping station, solvent dispenser, turbidity station, filtration station, pipette station, and MS/NMR racks.
The methods for sample preparation are described in Figure 1. Sample vials with bar-code labels are delivered to a sample chute on the system, which activates the system. The end result of the sample preparation method is the sample in solution is contained either in a capped NMR tube (5 mm X 8 in) with spinner or in a crimped MS autosampler vial (8mm conical, 0.8 ml), which is placed in an NMR or MS rack ready for transfer to the respective spectrometer. The preparation system uses 3 solvents for NMR samples (chloroform-d1, dimethylsulfoxide-d6, methanol-d4) and 1 solvent for MS samples (methanol).
Figure 1.
Sample Preparation Scheme for NMR and MS Samples
Instrumentation: A Varian Unity 400 NMR spectrometer equipped with a Zymark Zymate-XP robot is used to acquire NMR spectra on the samples prepared by the ORCA system. The Zymate-XP robot transfers the NMR tubes to the top of the upper barrel of the NMR probe from where the spectrometer takes over and acquires that data. Upon completion of the NMR experiment, the NMR tube is transferred back to the rack and the next tube is then introduced by the Zymate robot.
A Micromass Trio-2A mass spectrometer is used to acquire electron ionization (EI) and chemical ionization (CI) mass spectra from the rack of autosampler vials prepared by the ORCA robot. A GC autosampler transfers with a syringe 2 microliters of the sample solution to the direct exposure CI probe of the mass spectrometer. The solvent is allowed to evaporate with a stream of nitrogen gas, and the probe is controlled by the Trio-2 and driven by pneumatics. After passing through a vacuum lock, the probe is heated with an electrical current to desorb the sample. After cooling, the probe is removed using pneumatics and is ready for the next sample.
Computer Systems: The laboratory information management system (LIMS) is based upon the Beckman Lab Manager product (version 7.3) using their proprietary database structure and has been enhanced with several custom developed modules. The LIMS runs on a VAX 4000-300 computer using the OpenVMS operating system. There are 2 points of interaction between the LIMS and the laboratory automation. The first is during sample preparation when the ORCA system scans the bar-code on the sample vial, and then queries the LIMS for preparation conditions (i.e., NMR solvent). The second is when the spectrometer is being set up for a batch automation run. A worklist is downloaded from the LIMS to the spectrometer. The sample order is determined by the bar-code reading at the time of sample preparation. The worklist is formatted to be consistent with the sequence file used for executing batch analysis on the respective spectrometer.
The facility computer network is based upon ethernet and provides all the connectivity between spectrometers, robot systems, LIMS, departmental and desktop computers.
Spectrometer data files are stored on a VAX 4000-400 computer running OpenVMS and equipped with an NKK optical disk jukebox (35.8 GB capacity) and hard disk capacity of 7.1 GB. Data files are transferred from the spectrometer using PathWorks or FTP communication protocols. The user interface at the instruments is under Windows for the PC based mass spectrometers and under Open Windows for the UNIX based NMR spectrometer.
Results and Discussion
The sample preparation system for NMR and MS analysis was implemented in July 1994 and has prepared about 50,000 samples to date. The preparation time for NMR and MS samples is 4 minutes and 3 minutes, respectively. In a 24 hour day, this provides a theoretical capacity of about 400 samples. The practical limit is lower due to LIMS unavailability during backups and routine maintenance. Approximately 90% of the samples can be prepared successfully. The remaining 10% are prepared manually.
The automated NMR and MS spectrometers were implemented in 1992 and 1990, respectively. Each spectrometer generates between 10,000-15,000 spectra per year. The mass spectrometer was initially operated in a mode where EI mass spectra were generated first. If a molecular ion failed to be detected, the system would automatically run a CI mass spectrum in addition. As the emphasis with mass spectra has shifted toward confirmation of molecular weight, the MS system has been operating in CI mode only for the past year. The NMR spectrometer is operated in a variety of modes depending upon the request and requires only that the desired pulse sequence be programmed into the system in advance. Standard NMR experiments include proton, carbon, COSY, APT, HETCOR, fluorine and phosphorous. The cycle time required for generating a typical CI mass spectrum is 5 minutes. The cycle time for a standard proton NMR spectrum (1 mg of sample) is 7 minutes.
The LIMS plays a critical role in the automation of NMR and MS services. The LIMS supplies information on each sample to the preparation robot via a bar-code identification. The LIMS also tracks the sample order in the output rack of prepared samples and downloads the worklist to each spectrometer in the appropriate format for its sequence or run file. This eliminates redundancy of data entry and minimizes data entry errors and sample mix ups.
A custom sample login application was developed to simplify the process so that a casual user can perform the task with a minimum of effort. The login system was designed to be intuitive and handle transparently the necessary transactions with the LIMS. The login screen (Figure 2 and 3) was designed to resemble a paper submission sheet and prompt the user for needed information based upon their requests. This user login system allows samples to be entered into the LIMS when they arrive in the laboratory rather than at scheduled times. In turn, this allows the sample preparation tasks to be performed over a greater part of the day and frees up preparation capacity in the afternoon. This allows a cutoff time of 4:00 - 4:30 PM before the nightly run is initiated on the spectrometer. This system also eliminates the need for a part-time data entry specialist, which the previous system required.
Figure 2.
Sample Login Screen for Sample Identification Information
Figure 3.
Sample Login Screen for Available Analytical Services
The corporate intranet provides a powerful tool for providing information about the analytical chemistry laboratory. A web home page for the analytical chemistry laboratory has been created on the corporate intranet (Figure 4). This provides access to information on sample analyses that are completed or pending (Figure 5), on work completed on the sample preparation robot (Figure 6), on the sample queue for the open access NMR spectrometer, as well as various documents (SOPs and activity reports) and administrative functions.
Figure 4.
Web Home Page for Analytical Chemistry Laboratory
Figure 5.
Sample Status Report from Web Home Page
Figure 6.
Sample Queue on Open Access NMR System from Web Home Page
Spectral data files are stored on a VAX based optical disk jukebox system automatically at the same time results are posted to the LIMS. An instrument interface was developed that provides bi-directional data file transfer between the VAX and the spectrometer. This allows data files to be returned to the originating spectrometer for viewing or reprocessing. A user interface was developed for viewing and manipulating NMR spectra directly from the VAX based system. An X-windows application allows the database on the VAX based storage system to be queried by a number of parameters. The query results are seamlessly passed to an NMR data processing package for viewing and manipulation by the user.
Summary
The automation of sample preparation and instrument operation for NMR and MS have resulted in 1-1.5 scientists being able to be directed to higher level activities such as LC/MS experiments, advanced NMR experiments and to contribute to structure determination projects. This has eliminated a large number of manual and tedious tasks associated with sample preparation and instrument operation. In turn, this has provided an opportunity for the laboratory staff to learn about robotics, automation and computer technology. The user interface for sample login has allowed the customer to perform this task at any time during the day. This has provided greater utilization of the sample preparation system over the course of the day and freed up capacity during peak periods in the afternoon. It has also allowed the 0.5 person previously focusing on LIMS data entry to be directed toward scientific activities. Overall, the system handles 90% of all submitted samples without additional human intervention.
NetSci, ISSN 1092-7360, is published by Network Science Corporation. Except where expressly stated, content at this site is copyright (© 1995 - 2010) by Network Science Corporation and is for your personal use only. No redistribution is allowed without written permission from Network Science Corporation. This web site is managed by:
- Network Science Corporation
- 4411 Connecticut Avenue NW, STE 514
- Washington, DC 20008
- Tel: (828) 817-9811
- E-mail: TheEditors@netsci.org
- Website Hosted by Total Choice