Determining Protein Topology from Skeletons of Secondary Structures

1
Determining Protein Topology from Skeletons of
Secondary Structures

• Yinghao Wu, Mingzhi Chen, Mingyang Lu, Qinghua
  Wang and Jianpeng Ma
• Department of Bioengineering, Rice University

Journal Club 2005-10-17
2
Motivation

• Protein structures with low-intermediate
  resolution, such as those obtained by
  cryoelectron microscopy (cryo-EM)
• Computational method to interpret the detailed
  structural information

3
Formulate the Computational Question

• Given
• The skeletons of secondary structures derived
  from low-intermediate-resolution density map
• Protein secondary structure prediction from
  protein sequence
• Want
• Protein native topology/fold
• Secondary structure identity, direction,
  connectivity

4
Method - Flowchart

• Determine secondary structure skeleton
• Generate topology candidate
• Initial screening
• Geometry analysis
• Energetic analysis

5
Determine secondary structure skeleton

• Secondary structure skeleton in three-dimensional
  space
• Position, length, no connectivity
• Helixhunter, Sheetminer, Sheettracer
• Secondary structure assignment by primary
  sequence
• PSIPRED server
• If the two procedure disagree, 3D skeleton is
  more reliable than 1D assignment

6
Generate Topology Candidate

• Total number NT(Na!2Na)(Nß!2Nß)
• If mismatch

7
Initial Screening

• Length-based
• Loop length no longer than total extended loop
• Length variation 50 for alpha helices, 33 for
  beta strand
• At least two beta-strand in a beta sheet
• Sheet motif filter
• Knowledge-based

8
Geometry Analysis

• Packing geometry of two consecutive secondary
  structures
• Vector V1, V2, V3
• Angle ?1,?2,f
• Gaussian distribution
• Antiparallel, cis conformation

9
Geometric Analysis

• Geometry Scoring function
• Method II different loop motif

10
Energetic Analysis

• Hypothesis For a given protein skeleton, its
  native topology is chosen by evolution to
  accommodate the largest structural variation, not
  merely the one trapped in a deep, but narrow,
  energy well
• Construct structure ensemble for each topology
  candidate
• Secondary structure placement and loop
  construction by Monte Carlo (MC)
• Only C-alpha are considered in calculation

11
Energetic Analysis

• Loop Construction
• Within Loop, short-range bond energy
• Long-range interaction to avoid atom clash
• Additional energy between C- of Loop and N- of
  the next secondary element
• Global Optimization

12
Result All alpha helical proteins
13
Protein 1g7d
14
Result Sheet-containing Protein
15
Results

• Apply to incomplete skeletons
• Apply to real experimental data ?2 protein of
  reovirus

16
Conclusion

• A new energetic-based computational approach to
  derive protein topology from secondary-structure
  skeleton in space and assignment in sequence.
• Possible application to protein threading
  structural genomics