BMERC Seminar: Exploiting secondary binding pockets in aminergic GPCRs

June 21st, 2017

When:       Thursday, June 22, 12 pm

Where:     ENG room 203, 44 Cummington Street

Title:        Exploiting secondary binding pockets in aminergic GPCRs

Speaker:     Dr. Gyorgy Keseru, Medicinal Chemistry Research Group, Hungarian Academy of Science, Budapest, Hungary

Abstract:  Fragment based drug discovery (FBDD) employs growing and linking strategies for optimization. Structural information on G-protein coupled receptors (GPCRs) made FBDD available on this class of targets, however, most reported programs applied a growing strategy starting from orthosteric fragment binders. We developed a sequential docking methodology to support the identification of primary (orthosteric) and secondary site binders and linking of these fragment hits. Predicting the binding mode of multiple fragments bound to a single target we assessed the sampling and scoring accuracy for the first and second site binders in self- and cross-docking situations. The prospective validation of this approach was performed on dopamine receptors using the human dopamine D3 receptor crystal structure and a human dopamine D2 receptor homology model. Two focused fragment libraries were docked in the primary and secondary binding sites, and best fragment combinations were enumerated. Similar top scoring fragments were found for the primary site, while secondary site fragments were predicted to convey selectivity. A set of linked compounds created from the best scored primary and secondary site binders were synthesized from which we identified a number of D3 favoring compounds including one with 200-fold D3 selectivity. The structural assessment of the subtype selectivity of the compounds allowed us to identify further compounds with high affinity and improved selectivity. Now we are extending the methodology to further aminergic GPCR targets.

BMERC Seminar: Promiscuity in regulatory proteins such as PD-1 hinder their druggability

December 4th, 2016

When:       Thursday, December 8, 10:30 am

Where:      ENG room 203, 44 Cummington Street

Title:          Promiscuity in regulatory proteins such as PD-1 hinder their druggability

Speaker:  Prof. Carlos J. Camacho

Department of Computational and Systems Biology

University of Pittsburgh

Abstract:  Many regulatory proteins use structural flexibility to bind specifically to multiple partners. Although conformational selection and induced fit offer alternative pathways to rationalize multi-ligand binding, in practice we lack a structural understanding of how nature designs a flexible binding interface to select some ligands, but not others. Using molecular dynamics simulations (MDS) to identify the molecular interactions responsible for this “selective promiscuity”, I will show that promiscuity can be triggered by ligand-specific motifs that activate induced fit binding pathways toward different bound-like receptor states. This is the case for PD-1, a recent breakthrough anti-cancer immunotherapy target. Collectively, our studies provide insight pertinent to the structural basis and evolution of selective promiscuity in flexible regulatory proteins. Our findings also suggest a biophysical approach to exploit innate binding pathways to improve the druggability of seemingly undruggable targets.


BMERC Seminar: Targeting Protein-Protein Interactions and Surfaces of E3 Cullin RING Ubiquitin Ligases with Chemical Probes

September 17th, 2015

BMERC Seminar

When:       Monday, September 21, 2 pm

Where:      ENG room 203, 44 Cummington Street

Title:          Targeting Protein-Protein Interactions and Surfaces of E3 Cullin RING Ubiquitin Ligases with Chemical Probes: Small Molecule Inhibitors and PROTACs

Speaker:  Dr. Alessio Ciulli

Principal Investigator, Reader and BBSRC David Phillips Fellow at the School of Life Sciences, University of Dundee

Dr. Ciulli received his Ph.D. from Cambridge University, and conducted post-doctoral research there with Professors Chris Abell and Sir Tom Blundell on fragment-based and structure-based drug design. In 2013 he moved his laboratory to Dundee where he took up a Readership (Associate Professorship) in Chemical & Structural Biology. 

Abstract: This talk will focus on developing and characterizing small molecules binders of Cullin RING E3 Ubiquitin Ligases (CRLs) – the largest family of multisubunit ubiquitylating enzymes. CRL-targeting chemical probes can be used in their own right as E3 ligase inhibitors or modulators of the biological pathway in which the specific CRL is involved. In addition, CRL-targeting ligands can be suitably tethered with a ligand for a given protein of interest, yielding bifunctional proteolysis targeting chimeras (PROTACs) to hijack the ubiquitin proteasome system and induce the intracellular degradation of the target protein. I will describe recent progress from the lab with potent and selective VHL inhibitors and their applications in each of those areas, as well as current efforts at probing ligandability of CRLs using fragment-based and peptide-based approaches.


Galdeano C. et al. J. Med. Chem. 2014, 57 (20), 8657-8663.

Bulatov E., Ciulli A. Biochem. J., 2015, 467, 365-386.

Zengerle M. et al. ACS Chem. Biol., 2015, 10 (8), 1770-1777.

Seminar: Allostery and entropy in molecular recognition

May 26th, 2014

Our speaker had to cancel his visit this week.
We are very sorry for this change.
BMERC Seminar

When:       Thursday, May 29, 11:00 am

Where:      ENG room 203, 44 Cummington Street
Title:          Allostery and entropy in molecular recognition
Speaker:   Anatoly M. Ruvinsky
AstraZeneca, R&D Boston, 35 Gatehouse Drive, Waltham, MA


I will address two topics:1. Allosteric networks and iron pathways in eukaryotic and prokaryotic ferritins – Ferritin-like molecules show a remarkable combination of the evolutionary conserved activity of iron uptake and release that engage different pores in the conserved ferritin shell. It was hypothesized that pore selection and iron traffic depend on dynamic allostery with no conformational changes in the backbone. In order to detect a long-range network of cooperative residues, we have developed a network-weaving algorithm (NWA) that passes threads of an allosteric network through highly correlated residues using hierarchical clustering. The NWA was applied to four representative 24-meric ferritins ( BfrB, FtnA, bullfrog M and L Ftns) in conjunction with a packing-on elastic network model. We have detected the allosteric networks in all four ferritins and characterized their relationship to the iron pathways. The results show that the allosteric networks engage species-specific pores and a ferroxidase center in agreement with experimental observations of iron transport. The residue-residue correlations and the resultant structure of the allosteric networks depend significantly on the ferritin shell packing, which, in turn, depends on protein sequence composition. This relationship explains the diversity of iron pathways suggested by experimental approaches.2.  Accounting for binding entropy in protein-ligand docking – I will introduce several statistical-thermodynamic methods for calculating protein-ligand binding entropy in the context of predicting structures of protein-ligand complexes. These methods have shown substantial improvement in docking accuracy for AutoDock, D/F/G/X-Score, LigScore, PLP, LUDI, ChemScore, PMF, and DrugScore scoring functions. I will also discuss a relationship between binding entropy and a ligand propensity to bind in wide energy wells.


Dima Kozakov,

BMERC Seminar: New Directions in Docking and Scoring

May 23rd, 2014

When:       Tuesday, May 27, 11:00 am


Where:      ENG room 203, 44 Cummington Street


Title:          New Directions in Docking and Scoring


Speaker:   Dr. Istvan J. Enyedy
Biogen Idec, Cambridge, MA


Abstract: Target structure-based “hit” optimization in a drug discovery project is challenging from the computational point of view. Scoring functions cannot predict binding affinity, thus computational chemists must use their intuition or prior knowledge about the target class to prioritize compounds for synthesis. As the pharmaceutical industry targets novel protein classes, computational chemists must use software to build knowledge about the new targets. Over the last two years we have set up about 300 internal structures and more than 10,000 compounds for testing docking and scoring strategies. We used this set for evaluating the new scoring function CHEMGAUSS5, implemented in FRED and HYBRID from OpenEye, and ATLAS from Acpharis. The talk will focus on how we can guide docking and scoring by using solvent mapping for identifying and characterizing pockets suitable for small molecules. The use of fragment clusters for defining the shape and size of the search space used for docking or shape-based search will be shown. The presentation will highlight the influence of the binding site conformation, size of the search space, and target on the outcome of docking. In the end a comparison of statistical methods used for evaluating the performance of docking and scoring will be presented.

BMERC Seminar

April 1st, 2014

When:       Monday, April 7, 2:00 pm

Where:      ENG room 203, 44 Cummington Street

Title:          Interactive drug discovery technologies as complement to fragment based drug discovery

Speaker:  Prof. Carlos J. Camacho
Department of Computational and Systems Biology
University of Pittsburgh

Abstract:  Although there is no shortage of potential drug targets, there are only a handful of known low-molecular-weight inhibitors of protein-protein interactions (PPIs). One problem is that current efforts are dominated by low-yield high-throughput screening, whose rigid framework is not suitable for the diverse chemotypes present in PPIs. Here, we present AnchorQuery a novel pharmacophore-based interactive screening technology that builds on the role anchor residues, or deeply buried hot spots, have in PPIs, to increase hit rates by redesigning these entry points with anchor-biased virtual multicomponent reactions (MCR). This chemistry delivers hundreds of millions of readily synthesizable novel compounds especially suitable to disrupt PPIs, which typically are not amenable to traditional small molecule intervention. Our Google-like technology promises to expand the development of novel chemical probes for cancer research and the exploration of the human interactome by !
leveraging the designability of MCR to design inhibitors to one and/or multiple targets.

BMERC Seminar: Hybrid computational strategies for inhibitor design

February 10th, 2014

When:   Monday, February 24, 2:00 pm

Where:  ENG room 203, 44 Cummington Street 

Title:      Hybrid computational strategies for inhibitor design: a case study with protein methyltransferases

Speaker:  Melissa Landon, Ph.D., Schrodinger Inc., Cambridge, MA.

Abstract:  Protein methyltransferases (PMTs) are attractive epigenetic targets for the development of novel chemotherapeutic agents as well as for the treatment of neurogenerative- and inflammatory- disorders.  In this presentation I will discuss the application of hybrid computational strategies, i.e. combining both ligand- and structure-based approached, toward the identification of PMT inhibitors.  Furthermore, I will discuss a general framework for incorporation of 2D and 3D methods into a virtual screening campaign.

Bio: Melissa received bachelors’ degrees in chemistry and applied mathematics from Virginia Commonwealth University, and a PhD in Bioinformatics in 2007 from Boston University.  She completed postdoctoral training in x-ray crystallography in the joint laboratory of Greg Petsko and Dagmar Ringe before moving on to an industrial position at Cubist Pharmaceuticals, and to the Structural Genomics Consortium in Toronto, ON. Currently she is an Education Specialist with Schrodinger, a computational chemistry software company with offices in Cambridge, MA.

BMERC Seminar series

December 11th, 2013

Title: Occupy XLMS: Structural biology for the 99%

When: Friday , December 13th; 11:30-12:45am

Where: ERB705

Speaker:  Dr. Valdimir Svetlov (NYU)


Mass spectrometry-assisted mapping of covalently crosslinked peptides (XLMS) recently emerged as low-to-medium resolution method of structural analysis of proteins and their complexes. Here we present a robust workflow for XLMS analysis of bacterial transcriptional machinery and its structural bench-marking and validation. In vitro and in vivo cross-linking of proteins, computational discovery of cross-links and utilization of this information in structural proteomics will be discussed.

BMERC Seminar on peptide binding sites detection

July 24th, 2013

When: Friday, July 26, 3 pm

Where: ENG room 203, 44 Cummington Street

Title: Detection of peptide binding sites on protein surfaces

Speaker: Ora Schueler-Furman

The Hebrew University of Jerusalem

BMERC Seminar on Computational Protein Design

June 17th, 2013

When: Wednesday, June 19, 11 am

Where: ENG room 203, 44 Cummington Street

Title:    Conceptual framework for performing simultaneous fold and sequence optimization of reduced  representation protein models

Speaker: Istvan Kolossvary

D.E. Shaw Research, New York City, NY

Biographical Notes: István Kolossváry graduated with a Ph.D. and an M.Sc. in chemical engineering from the Budapest University of Technology. Prior to joining D.E. Shaw Research, István was an Associate Professor in the Chemistry Department at the Budapest University of Technology, and also served as a Visiting Professor at Columbia University. István was a Senior Research Scientist at the Novartis Institute for Biomedical Research, and made very important contributions to the methodology of conformational searches for large organic molecules. He also founded a research, software development, and scientific consulting company, BIOKOL Research, in 1999.

About D.E. Shaw Research: D.E. Shaw Research is engaged in scientific research in the field of computational biochemistry, including the design of novel algorithms and machine architectures for high-speed molecular dynamics (MD) simulations of proteins and other biological macromolecules. In particular, they have designed and constructed a specialized supercomputer called Anton, which executes such simulations orders of magnitude faster than was previously possible, along with a number of software tools and techniques that facilitate their execution and analysis.