Drug discovery, molecular biology and microbiology research depend upon robust libraries of small molecules. But libraries that catalog high-value compounds using controlled vocabularies and that evaluate chemical similarity using several different tools are, at best, difficult to find.
The Cheminformatics Shared Resource at UNM not only offers a library of over 4,000 small molecules but also routinely updates and adds to this body of information. In addition, the Cheminformatics group is available to help researchers narrow their search field to manageable amounts of work. By enabling focused and effective inquiry, this Shared Resource helps to move cancer research on difficult questions faster.
The UNM Cancer Center Shared Resources serve over 100 laboratories or principal investigators. They provide important technology support for our entire research community, both inside and outside the Cancer Center, across the HSC and UNM, and in association with our affiliated institutions. To begin collaboration, book time, or inquire about fees, please contact any of the Shared Resources listed at the end of the newsletter.
Scott A. Ness, PhD
The Victor and Ruby Hansen Surface Endowed Professor in Cancer Genomics
Associate Director, UNM Cancer Center
Professor, Internal Medicine
Director, Keck-UNM Genomics Resource
Cheminformatics Shared Resource
Focusing the Search
Finding a needle in a haystack is difficult at best, but finding a needle among twenty or so hay stalks is much simpler. The trick, of course, is narrowing the haystack to the smaller number of stalks.
For researchers wanting to look at large libraries of small molecules, the Cheminformatics group at the Flow Cytometry & High Throughput Screening Shared Resource can perform a similar feat. “We narrow the search space,” says Tudor Oprea, MD, PhD, Co-Director for Cheminformatics. And by narrowing the search space, they can reduce the workload.
The Cheminformatics group dwells at the interface between chemistry, physics, biology and medicine. “We try to offer a structured view of the drug discovery world, from a chemical perspective,” says Dr. Oprea. “And we tend to focus on small molecules, although biologics are becoming an area of interest.”
The small molecule drugs database that the Cheminformatics group maintains has over 4,000 molecules. About 1,600 of these are high-value compounds—ingredients of FDA approved drugs—with the remainder having been approved somewhere else in the world. These compounds have known toxicities and liabilities, so finding a new purpose for them could be quite lucrative.
One example of an FDA approved repurposed drug is ketorolac. Originally approved in 1991 for pain relief in humans, Angela Wandinger-Ness, PhD, and her team are now testing it in animal models and early phase human clinical trials for ovarian cancer use. Dr. Wandinger-Ness originally sought Dr. Oprea’s help to find a molecule that would control GTPases in a cancer cell. He identified R-naproxen from the initial tests. Further investigation revealed that R-ketorolac shares biologic activities with R-naproxen but unlike the former, R-ketorolac is approved for human use in the racemic form.
Reducing the search space is a blend of theoretical and experimental work. When the team of Eric Prossnitz, PhD, and Larry Sklar, PhD, were conducting work that led to the discovery of molecules to study GPR30, Dr. Oprea procured a library of 10,000 molecules from ChemDiv, which he used to search for GPR30 modulators. Given limitations at the time, Drs. Prossnitz and Sklar could not test all these molecules; in fact, they could test only a few hundred. So, Cristian Bologa, PhD, working with Dr. Oprea, used computational models and proposed only 100 molecules for testing . “We basically eliminated 99% of the library,” says Dr. Oprea, “and that reduced their workload significantly.” The result was the discovery of G1, a selective GPR30 agonist, followed by the discovery of G15 and G36, selective GPR30 antagonists, and an R01 grant funded by the National Cancer Institute.
Tools for Finding Patterns
To reduce a large library to a handful of compounds that warrant further investigation, the Cheminformatics group uses state-of-the-art technologies.
The controlled vocabularies that the Cheminformatics group implemented and maintains allows them to map the chemical structures of active pharmaceutical ingredients to diseases, off-label usage, and mechanisms of action. No publicly available resource affords database queries in this manner. The Cheminformatics group recently compared their library with a similar resource maintained by the NIH Center for Advancing Translational Sciences; discrepancies were few and minor, and over 90% of the UNM database was accurate. The use of controlled vocabularies permits multiple query modes, data mining and knowledge management, as annotations remain consistent across all drugs. So, the notion of “inflammation” is consistently mapped across all drugs, whether the search is for pharmacokinetic properties, indications, contraindications, off-label medical uses or targets.
The group also continually expands the library by data mining. Dr. Oprea, collaborating with Steven Seifert, MD, from the New Mexico Poison Control and Drug Information Center, published a paper in Clinical Pharmacology and Therapeutics in which they described the mechanism by which the muscle relaxant Flexeril (cyclobenzaprine) could cause the potentially lethal condition Serotonin syndrome. Cyclobenzaprine is very similar to the antidepressant Amitriptyline. In a series of experiments, Dr. Oprea, Dr. Seifert and their collaborators showed that cyclobenzaprine is active on some serotonin-related targets, similar to the antidepressant amitriptyline – which may explain Flexeril’s serotonin-related toxicity.
To find and evaluate molecular similarity, the Cheminformatics group uses a variety of methods. One such method describes chemical components of a molecule using a sequence of 1s and 0s, called a “binary fingerprint.” If the molecule contains a particular moiety such as an amide or a carboxyl group, then its bit string records a 1 in the location corresponding to that component, or zero if that moiety is absent. So, each molecule can be assigned a chemical signature, or a fingerprint.
To compare two molecules, the Cheminformatics group compares these fingerprints, or other related molecular descriptors, and evaluates the extent of their overlap. Identical bit strings imply perfect overlap, meaning that the molecules are very similar or the same. The more dissimilar two molecules are, the more dissimilar their fingerprints. However, these fingerprints indicate only presence or absence, not structure or properties.
“The other way to compare molecules is to think like a drug target,” says Dr. Oprea. “What they care about is shape and electrostatics.” The Cheminformatics group uses OpenEye® scientific software to rapidly compare the 3-dimensional shapes of molecules. Being able to “see” a molecule in space enables them to compare shape and other 3D-related properties for a large number of virtual or existing molecules, a process termed “virtual screening.”
These 3D methods can, for example, provide clues about where the molecule’s electrostatic potential regions are and what they look like. Using a target’s already-known 3-dimensional structure and its electrostatic characteristics, the Cheminformatics group can determine where a small molecule could fit, much like naval ships pulling into a specific port in a shipyard – hence the procedure is known as “molecular docking.” Knowing the shape and electrostatic characteristics of the docking region reveals what shape and electrostatic characteristics the docking molecule must have.
These tools to find small molecules are just that—tools. “There’s not a one-to-one relationship between what we compute and what’s out there in the real world,” explains Dr. Oprea. But by reducing the search space, the Cheminformatics group can help to focus research to the most potentially fruitful areas.
To learn more about the services and facilities the Cheminformatics group offers, please contact Dr. Tudor Oprea or visit the Flow Cytometry and High Throughput Screening website.
The Cheminformatics Team.
|The UNM Cancer Center supports and manages seven shared resources that provide services essential to basic and translational research. Each of these is available to researchers at UNM and affiliated institutions. Investigators do not have to be members of the UNM Cancer Center to use the facilities. Each of the resources is partially supported by user fees; use by Cancer Center members is subsidized through a cost-sharing mechanism. To learn more about specific services and how they might benefit your research, please refer to the service descriptions and contacts below.
Animal Models & Imaging Shared Resource
Helen Hathaway, PhD, Co-Director – Animal Models
Jeffrey Norenberg, PharmD, Co-Director – Animal Imaging – KUSAIR
Location: BMSB 155, NRPH B52 • Phone: 505-272-1469, 505-272-8242
Provides support for design, planning, and execution of in vivo assays through xenograft and genetically modified cancer models. The Keck-UNM Small-Animal Imaging Resource (KUSAIR) specializes in quanitative, in vivo molecular imaging using radiolabeled biomarkers and probes.
Christine Stidley, PhD, Co-Director for Biostatistics
Susan Atlas, Co-Director for Computational Biology
Location: Research Incubator Building (RIB), Suite 190 • Phone: 505-272-2520, 505-272-8704
and Center for Advanced Research Computing, UNM Main Campus.
Offers biostatistical collaboration and support for study design, data analysis, clinical trials, grant preparation and methodological development. Provides expertise in the design and analysis of genetic studies, including next-generation sequencing, methylation and gene expression research.
Teresa Stewart, Administrative Director
Melanie Royce, MD, PhD
Zoneddy Dayao, MD
Location: UNM Cancer Center, Administration Bldg., 2nd floor • Phone: 505-272-5490
Administers all aspects of the clinical trials process from implementation of new trials to data safety monitoring and quality assurance.
Bruce Edwards, PhD, Director
Tudor Oprea, MD, PhD, Co-Director for Cheminformatics
Location: Innovation, Discovery & Training Complex, Bldg. 289
Phone: 505-272-6206, 505-272-8223
Provides instrumentation and expertise for fluorescent cell and particle analysis and sorting plus access to high throughput drug screening, including a wide-range of screening and drug informatics approaches focused on systems chemical biology, pathway analyses, patent analytics and chemoinformatics. View the Center for Molecular Discovery HSC Youtube video.
Angela Wandinger-Ness, PhD, Director
Location: CRF 212, 216, 218 • Phone: 505-272-5876
Provides instrumentation and expertise for analysis of cellular and subcellular structures using conventional, confocal and two-photon fluorescence microscopy approaches.
Therese Bocklage, MD, Medical Director
I-Ming Chen, DVM, Resource Scientific Director
Kelly Higgins, PhD, HTR Senior Operations Manager
Cathleen Martinez, HTL, PA, HTR Technical Manager
Location: BMSB 306 C • Phone: 505-272-1127
Collects, maintains, prepares and distributes patient samples (e.g., blood, sputum) and normal and tumor tissue and provides advanced histology services for research use.
Scott Ness, PhD, Director
Jeremy Edwards, PhD, Co-Director
Location: CRF 118 • Phone: 505-272-5564
Provides expertise and instrumentation for generating and analyzing the data from all types of gene expression and genomic analysis including Ion Proton next-generation sequencing for exome, RNA-seq and epigenetics assays, Affymetrix microarrays and real-time PCR.
The UNM Cancer Center is a leader in cancer research and treatment. One of just 68 National Cancer Institute-designated cancer centers in the nation and the only such center in New Mexico, the UNM Cancer Center is recognized for its scientific excellence, contributions to cancer research and delivery of medical advances to patients and their families. It is home to New Mexico's largest team of board-certified oncology physicians representing every cancer specialty and 126 cancer research scientists. The Center’s research programs are currently supported by over $72 million annually in federal and private funding.