The Family Business

Computational Biologist David Baker:

Modeling Protein Structures

---

David Baker Shapes Proteins

Computational biologist Baker is fighting disease with a better picture of proteins.

Proteins

• Consist of amino acids

• Attach to other molecules

• Trigger chemical reactions in the body

Question:

How do scientists predict protein shapes?

---

Answer: Using approaches from physics and computer sciences

---

Rosetta Software Models Protein Structure

---

Making a Protein from Scratch

---

Many Rosetta Flavors

Different versions of Rosetta

• Predict how a protein interacts with another protein or with DNA

• Add in experimental data

• Use structural information of other, similar proteins

---

Develop new drugs and vaccines based on interactions

Create custom proteins to interrupt or enhance reactions inside a cell

Goals of Computational Biology

Track interactions

of single proteins with other molecules

Develop accurate models of protein structures

Identify atoms, bonds, and places where chemical reactions occur

Computational biologists have more than one goal

---

Challenges & Solutions

• Insufficient computing power

• Simulations of 10,000 possibilities can take 100 days

• Resolution of models

• Low resolution models are not 100% accurate

---

Community-wide Experiment

Critical Assessment of Techniques for Protein Structure Prediction (CASP)

• Scientific competition

• Entrants from >200 labs worldwide

• Work together

• Learn about current challenges

• Set future goals

• Assess methods and technology for predicting protein structures

---

Team Baker's CASP Rainbow

Protein's actual X-ray crystallographic structure

Team's computer model

Rosetta highlighted even more detail than X-ray method

---

Research Applications

How might custom proteins keep people healthy?