You are here

Rosetta Commons Summer Internships

Rosetta Commons Summer Internships

The Rosetta Commons is pleased to be offering undergraduate student internships for the summer of 2018.

The program:

  • One week of Rosetta Code School (June 4 through June 8) where you will learn the inner details of the  ​Rosetta C++ code and community coding environment, so you are fully prepared for the summer!
  • 8 weeks of hands-on research in a molecular modeling and design laboratory, developing new algorithms and discovering new science.
  • The summer will finish with a trip to the Rosetta Conference in the gorgeous Cascade Mountains of Washington State, where you will present your research in a poster and connect with Rosetta developers from around the world. The conference will be held from August 6 through August 10. 
  • This program is supported by NSF. Interns will receive housing, travel expenses up to \$500, and a stipend of \$5,500.



  • College Sophomores or Juniors preferred
  • Major in computer science, engineering, mathematics, chemistry, biology, and/or biophysics
  • Available for at least 10 weeks during the summer of 2018. 
  • Interest in graduate school
  • U.S.citizens, permanent residents, and U.S. nationals are eligible.  
  • While not required, we seek candidates with some combination of experiences in scientific or academic research, C++/Python/*nix/databases, software engineering, object-oriented programming, and/or collaborative development (git).


To apply:

  • Complete and submit the application, open in November 2017. 

  • Include the following in the application:
    • Resume
    • Transcript
    • Personal statement - why this internship interests you - brief summary of research and computing experience - why you are a good candidate for the position (up to 4000 characters)
    • Two references (complete the reference forms with contact information)
  • Deadline for receipt of applications and recommendation letters is February 1, 2018.
  • For questions, please contact Camille Mathis at


Available projects and locations:

Baker Lab @ University of Washington in Seattle, Washington

“Design of self-assembling protein nanomaterials: three-dimensional crystals”

Designed protein crystals have great potential as structure determination tools and as biomaterials for energy, environmental, and health-related applications. While many proteins will crystallize under artificial vapor diffusion conditions, no method currently exists for controlling the crystallization of proteins into specific lattice architectures. The general rules for protein crystallization can be broken down into chemical, physical and geometric constraints. Incorporating these parameters into tools for the accurate structural modeling of macromolecules allows for the design of proteins. The project will use Rosetta modeling to design self-assembling protein crystals constructed from complex protein scaffolds with atomic-level accuracy.


Cooper Lab @ Northeastern University in Boston, Massachusetts

“Crowdsourcing Protein Folding and Design”

We are looking at how crowdsourcing can used be help biochemists with their work. To do this, we have developed the game Foldit. Foldit is a multiplayer online game that allows players without previous experience in biochemistry to work on protein folding and design problems. The players interactively manipulate protein structures to find structures with good energies. Tutorial levels introduce the gameplay and weekly challenges allow players to compete and collaborate. This project will be primarily focused on development of game-related aspects of the project and improving the player experience.
(C++ or other coding skills, game design/development/UX experience or interest would be great) 



Correia Lab @ Swiss Federal Institute of Lausanne(EPFL) in Lausanne, Switzerland                                                                                                       

“Design of epitope-focused immunogens for ricin vaccine development”     

Despite the generalized use of vaccines to prevent viral infections, the broad scope of activities exerted by antibodies enables the utilization of vaccination for a plethora of pathological agents, such as toxins, cancer-related malignancies and others. Here, we propose to explore new computational technologies of immunogen design applied to the development of a ricin vaccine. Ricin is a toxin, which is considered a potential bioterrorism agent and poses relevant threats to our safety and wellbeing.                          

Protective levels of neutralizing antibodies are generally recognized as a required component of efficacious vaccines. To leverage this observation, a new class of epitope-focused immunogens has emerged from a vaccine development strategy dubbed reverse vaccinology. This strategy relies on the identification of neutralizing antibodies and structural characterization of their epitopes, followed by immunogen engineering for epitope stabilization and presentation. These optimized immunogens aim to be superior neutralizing antibodies inducers and the foundation of epitope-focused vaccines.

The development of ricin vaccine stands out as an important goal, here we will apply state-of- the-art techniques to design optimized immunogens that will aim to improve disadvantages shown by ricin protein’s based immunogens, such as thermodynamic instability and the elicitation of toxicity enhancing antibodies – two major issues for vaccine immunogens. Our study on ricin vaccine development can contribute to forge a roadmap on how to develop epitope-focused immunogens for other relevant pathogens in urgent need of an efficacious prophylactic vaccine (e.g. HIV, Flu and others).


Fomekong-Nanfack Group @ EMD Serono Research and Design Institute, Inc. in Billerica, Massachusetts.

“Docking, Prediction & Design using Rosetta to enhance the drug discovery process”

The Drug Structure, Prediction, and Design (DSPD) group is an interdisciplinary team of scientists built around a core of experts to drive drug discovery of small molecules and biologics. We blend experimental science with quantitative and qualitative in silico approaches that span computational chemistry and biology, structural biology, mathematics and data science.

A Rosetta intern in DSPD will choose from one of three project ideas that will support the Discovery Technologies drug candidate portfolio by helping drive an understanding what makes a molecule a good drug candidate. The projects include 1) docking experiments to recapitulate or predict small molecule-protein ternary complexes for the immune modulatory (IMiD) class of drugs; 2) modeling of full antibody molecules to predict self-association and derive predictive models for aggregation; and 3) Designing antibodies to bind to cancer specific peptides.

Each of these has the potential to significantly enhance the drug discovery process through computational means.


Gray Lab @ Johns Hopkins University in Baltimore, Maryland

“Glycoengineering for immunotherapy and biofuels”

Glycosylation is a key marker of cancer cells, and biofuels are comprised of cellulosic material.  Both require modeling of the structure and dynamics of the underlying sugars comprising the molecules.  This project will apply new glycoprotein modeling tools for these critical applications.


Jiang Lab @ University of California, Los Angeles in Los Angeles, California

"A hybrid, integrative approach to build high-resolution structural models of amyloid protein aggregates in neurodegenerative diseases"

Our limited knowledge of the atomic-resolution details of amyloid structures impedes the development of therapeutics to treat amyloid diseases, including Alzheimer’s disease.  The goal is to enhance our structural knowledge through experimentally informed computational modeling. We plan to incorporate available structural data to guide and constrain the model construction of atomic structures of both amyloid protein fibrils and oligomers.  These detailed models will make it possible to connect sequence data, including familial disease mutations, to function data, such as interaction partners and toxicity, and will thereby offer a better understanding of the role of amyloid aggregation in disease development.  This structure-based analyses using Rosetta symmetry modelling will bridge the current gap between genetic discoveries from genomic research and molecular mechanisms leading to disease, revealing novel targets for therapeutic intervention.We will apply a hybrid, integrative approach to make testable structural models to integrate, explain, and guide experiments for α-synuclein fibrils in Parkinson's disease, Aβ fibrils and oligomers of Alzheimer’s disease, Tau fibrils associated with various neurodegenerative disorders and prion protein fibrils in prion transmission.


Kortemme Lab @ University of California, San Francisco in San Francisco, California

 “Computational design of protein-based sensor/actuators”

Detecting signals (sensing) and responding to them (actuating) are among the most fundamental abilities of living systems. Natural proteins that sense and respond to small molecule signals can regulate very complex biological behaviors, such as production of chemicals in metabolism, formation of microbial communities in the human gut, or differentiation of cells in mammalian tissues. Our goal is to engineer new sensor/actuators that do not currently exist in nature. Such new functions would be useful in many applications, including manufacturing valuable chemicals in microbes, building new communities of cells, and probing protein-protein interactions in signaling networks that control inflammation and cancer. Very recently, we have engineered a completely new sensor/actuator that functions in living cells and detects a key intermediate on pathways leading to diverse compounds, including a malaria drug and other value-added chemicals. We determined the three-dimensional structure of the sensor, which was very close to our Rosetta model. Our REU project will be based on this excited result and will (i) develop and test Rosetta methods to improve new sensors we have designed, and (ii) modify their specificity to detect related pathway intermediates. Ultimately, we plan to apply these methods to design many other new sensors.


Kuhlman Lab @ University of North Carolina in Chapel Hill, North Carolina

“Iterative Design of Hydrogen Bond Networks”

Hydrogen bonds provide both stability and specificity to macromolecular interactions, and play a critical role in the formation of protein-protein complexes.  Our laboratory develops methods for computational protein design, and has a particular interest in designing protein-protein interfaces.  This problem is particularly challenging because the position and conformations of the binding partners have to be established in such a way that the amino acids at the interface can form tight van der Waals interactions, favorable electrostatic interactions, and low energy hydrogen bonds.  Forming extensive hydrogen bond networks in which each bond has a favorable geometry is challenging because the optimization/design of one bond often influences the positioning of neighboring bonds.  Here, we aim to develop an iterative protocol for designing hydrogen bond networks that starts with design of a few interactions, and then perturbs the system with gentle rocking or backbone motions in search of additional conformations and side chain positions that can extend the network.  We anticipate that this will be a computationally intensive problem, and therefore will develop a new job distribution system in Rosetta that allows for adaptive sampling, and thus can be used to redirect trajectories mid-protocol towards more productive regions of conformational and sequence space.  The end result of the project will be a new protocol for redesigning the surface of one protein so that it will bind to pre-specified target proteins.  These binders will be useful as competitive inhibitors and biosensors, and therefore can be used to probe signal transduction in living systems.  


Meiler Lab @ Vanderbilt University in Nashville, Tennessee

“Rosetta Computer Aided Drug Design”

The filovirus family includes Marburg and Ebola viruses, most of which cause highly lethal hemorrhagic fever. The first filovirus was identified when it sickened laboratory workers in Europe in 1967. Since then, filo-viruses have re-emerged multiple times, with modern strains conferring greater lethality (~90%). Ebola virus is typically found in Central Africa, but re-emerged in Western Africa in 2014 to cause an on-going outbreak unprecedented in magnitude and geographic spread that has already claimed the lives of thousands of people.

Early 2015 structural data on several neutralizing antibodies (Abs) in complex with the viral target envelope surface glycoprotein (GP) became available to us, enabling for the first time the research proposed within this PAPID proposal. Specifically, co-crystal structures of one Ab with the Marburg and Ebola GP as well as electron microscopy (EM) density maps for six additional Abs. It is our Goal to investigate the shared structural determinants of human Abs neutralizing the filovirus. While the present study thereby focuses on the biological question of the immune response to an infection by the Ebola and Marburg viruses, it has the potential to facilitate development of therapeutic strategies by others.

Within the scope of project we pursue three goals: Aim 1 creates atomic detail models of the Ab/GP interface for all seven Abs with available EM density maps using PHENIX.ROSETTA5 and ROSETTA refinement into density maps. We will use the co-crystal structure to test our computational protocol, and then apply it to the remaining six cases. As a result, we will obtain a map of critical structural determinants an Ab needs to fulfill for being neutralizing. Aim 2 will identify novel Abs from the repertoire of naïve subjects that are likely to neutralize Ebola and/or Marburg viruses with a limited set of mutations. We also will compare the Ab repertoires of filovirus-naïve (i.e. not previously infected) humans to those of immune (previously infected) subjects available to us. These Abs from naïve subjects are important, as they could evolve into neutralizing Abs upon infection or vaccination. We will redesign these Abs computationally to bind and neutralize Ebola and/or Marburg viruses, to increase the pool of known neutralizing Abs, paving the way for a successful vaccination strategy. In Aim 3 we will characterize these Abs experimentally for binding and neutralization activity.


Mills Lab @ Arizona State University in Tempe, Arizona

“Design of proteins using expanded genetic codes”

In the last 15 years, the genetic codes of organisms including bacteria, fungi, and even mice have been expanded to include over 150 amino acids that are not found in naturally occurring proteins. Our lab seeks to ask what proteins might look like if such unnatural amino acids had been present during evolutionary history. To explore this question, we combine Rosetta’s computational methods with experimental characterization of the designed proteins to engineer functional proteins that can be used to study biological systems, sense foreign molecules, and treat diseases.

Professor Jeremy Mills has had a longstanding interest in education and scientific outreach in university, local school, and local community settings. In addition to judging science fairs at the regional, state, and international levels in Tennessee, Texas, California, Washington, and Arizona, Jeremy served as an inaugural lecturer at HiveBio, a DIY community laboratory in Seattle. In the HiveBio lectures, the online protein design game Foldit was used as a platform to teach community members fundamental concepts of protein structure and function. Professor Mills has also had a longstanding interest in fostering undergraduate education outside of the classroom and has served as an advisor of the UW and ASU iGEM teams.


Sgourakis Lab @ University of California Santa Cruz in Santa Cruz, California

“Modeling macromolecular assemblies from sparse NMR data”

Our research focuses on elucidating the ground-state structures and dynamic transitions of important macromolecular complexes involved in immune recognition of viruses and tumors, bacterial secretion and neurodegeneration.

We strive to predict the emerging properties of biological systems from their intrinsic self-assembly behavior. To do this, we are developing and implementing new hybrid structural biology tools using a range of experimental solution techniques (NMR, SANS/SAXS) alongside advanced computational sampling methods. The integration of experimental and computational approaches enables structural studies of protein complexes at high resolution.


Siegel Lab @ University of California, Davis in Davis, California

Our long term goal is to develop a suite of tools that enable on-demand access to any physically possible biochemical transformation of interest. In order to achieve this goal, we must be able to rapidly and reliably identify enzymes with desired functions. To this end we have recently laid the foundation for a platform that will enable a community of scientists to be trained and carry out enzyme engineering efforts in order to work together towards a common goal of understanding enzyme sequence, structure, and function relationships. We are calling the platform SEEK: Standardized Enzyme Engineering Knowledgebase (

SEEK provides methods and materials for a complete Design-Build-Test-Learn cycle for enzyme engineering, as well as data sets derived from those methods. The first stage in the cycle is computational modeling and design of an enzyme, for which the program Foldit, template files, and instructions for use are provided (Design). Once modeling is complete, detailed protocols are provided to readily and reliably build the designed enzymes (Build). Additional files and scripts are provided to guide students through enzyme assays and data analysis (Test). With data in hand, students can use previously reported data sets, readily available on SEEK, to compare with their results using a combination of modeling and machine-learning algorithms in order to develop new data-driven hypotheses explaining their physical observations (Learn). Ultimately, we aim for SEEK to become a comprehensive, and easy to access, database of kinetic constants derived from enzyme engineering through the SEEK Platform.


Whitehead Lab @ Michigan State University in East Lansing, Michigan

”Protein engineering to improve synthetic metabolic pathways”

Synthetic metabolic pathways can now be reconstituted into chassis organisms like Baker’s yeast to produce next generation fuels, complicated pharmaceuticals, and value-added commodity chemicals. However, the enzymes comprising these pathways are often unstable in their new environment. Our group uses massively parallel experiments to experimentally evaluate the effect on enzyme stability and solubility of each point mutation in a protein sequence. However, most mutations imparting stability identified by this screen decrease the catalytic efficiency of the enzyme. The goal of this project is to develop a filtering mechanism using Rosetta macromolecular software package to identify combinations of mutations likely to result in active, more stable enzymes. This computational screen to be developed will comprise homology modeling, phylogenetic relationships, and all-atom simulations. The student will engage in both computational design and experimental characterization of her designs.



The Intern Experience:


Intern Research Posters: