Research Aide - DSL - Cao, Adrian - 4.9.26.
About The Position
Intrinsically disordered proteins (IDPs) are therapeutically important but notoriously difficult to drug because they lack stable three-dimensional structures, meaning no single binder can engage the full conformational ensemble. We are experimenting with a hierarchical swarm of AI agents that collectively tile the protein-protein interaction (PPI) surface of an IDP target by deploying specialized binder design tools across distinct epitope subspaces and scaffold topologies in parallel on the Aurora exascale supercomputer. Several analytical and coordination tools within the swarm are not pre-built but are instead generated on-the-fly by the reasoning models themselves through code generation, allowing the system to synthesize custom analysis pipelines — such as novel residue foot printing routines or binder scoring functions — in direct response to emerging experimental findings without requiring manual software development. The swarm is coordinated through a shared blackboard that maintains a five-objective Pareto archive balancing binding free energy, PPI proximity, scaffold developability, epitope coverage, and structural diversity, with frontier reasoning models like GPT-OSS-120B embedded as first-class computational components performing cross-generation hypothesis synthesis, adaptive hyperparameter governance, and cooperative/competitive binder relationship classification. The multi-objective decomposition is handled via MOEA/D, where each swarm agent owns a scalarized subproblem over a distinct region of objective space, and a suite of blackboard reasoning agents adaptively redistribute weight vectors, detect hotspot-redundant candidates via MinHash LSH, and prevent scaffold mode collapse through a closed-loop diversity controller. The net result is a system capable of autonomously discovering a diverse panel of IDP-targeting biologics that collectively disrupt a target's full PPI network — a qualitatively richer output than any single-objective or single-agent binder design approach can produce, and a concrete demonstration of autonomous scientific reasoning at exascale.
Requirements
- The entirety of the appointment must be conducted within the United States.
- Applicants must be: o Currently enrolled in undergraduate or graduate studies at an accredited institution. o Graduated from an accredited institution within the past 3 months; or o Actively enrolled in a graduate program at an accredited institution.
- Must be 18 years or older at the time the appointment begins.
- Must possess a cumulative GPA of 3.0 on a 4.0 scale.
- If accepting an offer, candidates may be required to complete pre-employment drug testing based on appointment length. All students remain subject to applicable drug testing policies.
- Must complete a satisfactory background check.
Stand Out From the Crowd
Upload your resume and get instant feedback on how well it matches this job.
What This Job Offers
Job Type
Full-time
Career Level
Intern
Education Level
No Education Listed
Number of Employees
1,001-5,000 employees
