The Computational Structural Biology Section studies structural mechanisms of allosteric drug action, formulates determinants of oncogenic cell signaling, and designs agonist/antagonist drugs. 


Pioneering dynamic free energy landscape and cell function 

We draw on multidisciplinary advances to decipher the mechanisms and pharmacology of oncogenic proteins and their signaling at the molecular level. Our work is based on our concept of dynamic free energy landscapes. The landscape describes protein molecules as consisting of preexisting ensembles of conformations, whose relative populations determine function.   

We pioneered proteins existing as dynamic conformational ensembles whose propensities can be predictors of cell function. We proposed that the conformational states that the molecules preferentially visit can be viewed as phenotypic determinants and that their mutations work by altering the relative propensities, thus affecting the cell phenotype. Our team further postulated the concept of conformational selection and population shift as an alternative to the induced-fit textbook model to explain molecular recognition and allosteric regulation. Further experiments validated these ideas.  

We extended the ensemble model to catalysis, oncogenic activation, and mechanisms of inhibition, contributing to substantial advances in the scientific community's understanding of structure, function, and gain-of-function. Our approach clarifies observations and makes predictions, aiming to deepen understanding about experimental and clinical observations.   

RAS proteins and downstream pathways 

The section focuses on RAS proteins; their oncogenic activators, including tyrosine kinases; and downstream pathways, including PI3K, PTEN, and MAPK (B-Raf/KSR/MEK).  

RAS proteins and networks are important in multiple biological contexts, including cancer. Our scientists concentrate on four major aims:  

  1. Uncovering the mechanisms of activation of key nodes in cellular pathways, how their cancer-activating mutations work, how biological processes can inhibit them and are relieved, and how the emitted signals propagate to drive cell proliferation 
  2. Discovering what determines the signaling strength, since many of the cellular requirements haven't yet been considered or resolved 
  3. Harnessing innovative concepts to address questions like why cancers with a common driver mutation aren’t likely to evolve a common drug resistance mechanism, or whether we can predict the likely mechanisms that the tumor cell may develop 
  4. Discerning mutations, pathways, and mechanisms in cancer versus neurodevelopmental disorders 


Additional Content

Unraveling the KRAS signaling network’s cancer-promoting mechanisms 

The Computational Structural Biology Section is working to uncover the mechanisms by which the KRAS signaling network drives cancer at the conformational and cellular levels. The first, which involves studying proteins' conformational dynamics, propensities, and interactions, is vital for defining the mechanisms involved in the process and for understanding direct, same-protein drug combinations. The second is critical for insight into homeostatic mechanisms, including feedback control, bypassing, and complementary pathways. Together, these are vital for designing drug combinations.  

Our capabilities and specializations

Additional Content

Investigating protein structure, regulation, and signaling 

Using computational and experimental approaches, we investigate protein structures, their dynamic conformational ensembles, and interactions to advance biological and clinical applications. We elucidate how same-allele mutations in oncogenic proteins can promote cancer and other diseases, such as neurodevelopmental disorders.

Additional Content
  • Offer innovative concepts for how protein mutations provoke disease  

  • Evaluate data to expedite drug discovery and development 

  • Model conformations in membranes, their activation, and inhibition 

  • Explore protein structure and dynamic behavior to decipher physiologic and oncogenic activation mechanisms and signaling 

  • Decipher the physical and chemical behavior of molecules in cells 

  • Predict protein–protein interactions between pathogen and human host proteins to explore pathogenesis and inhibition  

  • Construct a server for the scientific community for predicting human–pathogen protein–protein interactions 

Additional Content

Decoding cancer drivers 

A key focus for the sections involves investigating KRAS proteins and how they can drive cancer. We work to decipher mechanisms in the activation and signaling of oncogenic KRAS proteins and key proteins in their major signaling pathways, and we study the signaling behavior of RAS protein variants at the cell membrane. 

Additional Content
  • Decipher activation mechanisms of KRAS proteins regulators at the membrane 

  • Determine the activation mechanisms of oncogenic PI3K, B-Raf, and PTEN and other key proteins in oncogenic RAS signaling and their complexes (e.g., B-Raf/MEK and B-Raf/KSR) 

  • Forecast potential parallel signaling pathways and key proteins that can be targeted to mitigate drug resistance  

  • Resolve apparent experimental contradictions in RAS signaling 

  • Investigate how wild-type KRAS proteins can inhibit their mutated cancerous variants 

Additional Content

Devising a structure-based platform for precision medicine and emerging treatments 

The section utilizes a multipronged approach to drug discovery, combining computational data, clinical observations, experimental approaches, genetics, biophysics, and protein structure data to alleviate treatment limitations. We consider the availability of each protein in each pathway in the specific cell, its activating mutations, and the chromatin accessibility of its encoding gene.  

Additional Content
  • Capitalize on new, protein-structure-based concepts to develop and integrate computational algorithms and experimental approaches 

  • Combine statistical analysis with a genetic and molecular structural basis to improve cancer treatment decisions 

  • Integrate computational methods, functional assays, and conformational principles for interpreting cancer drivers  

  • Develop a powerful deep learning computational methodology to accelerate the identification of novel drug–target interactions