Degree Type


Date of Award


Degree Name

Doctor of Philosophy


Physics and Astronomy


Bioinformatics and Computational Biology

First Advisor

Alex Travesset

Second Advisor

Guang Song


RAS subfamily proteins regulates cell growth and differentiation by cycling between active (GTP-bound) and inactive (GDP-bound) states. Mutations intervening normal Ras functioning are associated with several human cancers and developmental disorders. The three RAS isoforms in human HRAS, KRAS, and NRAS are the most common oncogene found in human cancers. Despite considerable experimental and computational efforts, it has remained difficult to achieve a therapeutic grip on RAS proteins mainly due to the incomplete understanding of the intermediate structures in RAS GDP-GTP transitions. Identifying the distinct features of intermediate states in RAS signaling processes is thus highly desirable for the design of small molecule inhibitors. The primary focus of this work was to develop a generic coarse-grained model of proteins, use it to study conformational transitions in RAS proteins with the goal to identify the critical structural features controlling the intrinsic conformational transitions, and complement the results using all-atom simulations. Knowledge of such critical features is bound to provide invaluable understanding of the ways in which these processes would be catalyzed by regulatory proteins. Thus, the present work also lays the foundation for future works involving coarse-grained modeling of RAS conformational switch mechanisms in the presence of regulatory proteins.

In the first part of the thesis, we developed a coarse-grained model that successfully folded nineteen different proteins into their native states (containing β-sheet, α-helix, and mixed α/β) starting from completely random configurations. The model is sensitive to small changes in protein sequence, and more importantly, the results obtained from the coarse-grained model were shown to complement very well with results from all-atom molecular dynamics. Using coarse-grained simulations in combination with all-atom simulations (total of 3.02µs) of HRAS, we identified the structural features that regulate the intrinsic nucleotide (GDP) exchange reaction. Our results suggests that dissociation of GDP/Mg from the nucleotide binding pocket is initiated by a loss of interaction between GDP and the base binding region of RAS. Further, we provide the first simulation study showing displacement of GDP/Mg away from the nucleotide pocket in both mutant and wild-type RAS. Both SwitchI and SwitchII, the known critical elements in RAS signaling, delay the escape of displaced GDP/Mg in the absence of guanine nucleotide exchange factors (GEFs). A model for the mechanism of GEF in accelerating the exchange process is presented. We also provided a comprehensive comparison of the dynamics of all the three RAS isoforms using extensive molecular dynamics simulations in both the GDP- (total of 3.06µs) and GTP-bound (total of 2.4µs) states. We identified a new pocket on RAS structure, which opens transiently during MD simulations, and can be targeted to regulate the nucleotide exchange reaction or possibly interfere with membrane localization. Furthermore, we have identified a new cluster of wild-type GTP-bound structures that potentially represents an intermediate conformation in the GTP hydrolysis process.

Copyright Owner

Abhijeet Kapoor



File Format


File Size

238 pages