Amarda Shehu
Papers
15
Total Citations
311
H-Index
11
About
Amarda Shehu is a computational biologist and computer scientist whose research sits at the intersection of robotics-inspired algorithms, machine learning, and structural biology. Her work centers on two deeply interconnected challenges: ab initio protein structure prediction and the modeling of protein conformational dynamics and transitions. Shehu's most influential contributions involve adapting probabilistic roadmap and tree-based exploration methods—originally developed for robotic motion planning—to navigate the astronomically complex conformational spaces of proteins and peptides. Her 2010 paper on tree-based exploration of native-like protein conformations (56 citations) demonstrated that robotics-inspired search strategies could meaningfully guide structure prediction using only amino acid sequence information. Subsequent work refined these approaches through energy-guided probabilistic sampling and structural profiles, tackling the persistent challenge of decoy generation in template-free prediction. Beyond static structure prediction, Shehu has made notable strides in characterizing protein transition pathways between functionally-relevant states, recognizing proteins as dynamic molecular machines rather than rigid structures. Her 2016 survey of robotics-inspired computational treatments of biomolecules (22 citations) synthesized decades of progress in this emerging interdisciplinary field. Collectively, her papers have garnered hundreds of citations, establishing her as a leading voice in applying algorithmic and probabilistic thinking to fundamental problems in structural and functional genomics.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4
- 5
- 6
- 7
- 8
- 9Sampling Conformation Space to Model Equilibrium Fluctuations in Proteins17 citations · 2007
- 10A General, Adaptive, Roadmap-Based Algorithm for Protein Motion Computation15 citations · 2016