Hydrogen bonds in Rosetta: a phenomonological study

1
Hydrogen bonds in Rosettaa phenomonological
study

• Jack Snoeyink
• Dept. of Computer Science
• UNC Chapel Hill

2
Key points

• My biases
• Hydrogen bonds in Rosetta
• Structure-derived potential of KMB03
• Existing definition/scoring
• Comparing natives decoys
• Proposed recategorization
• Bad smells in code
• Open questions

3
Phenomenology defined

• Movement originated by E. Husserl in 1905
• A philosophy based on the premise that reality
  consists of objects and events as they are
  perceived or understood in human
  consciousness and not of anything independent
  of human consciousness.

4
Phenomenology defined

• Movement originated by E. Husserl in 1905
• A philosophy based on the premise that reality
  consists of objects and events as they are
  perceived or understood in human
  consciousness and not of anything independent
  of human consciousness.

5
Structure-derived potential KMB03

• Energy from observed structures distance
  dependence for helix

6
Structure-derived potential KMB03

• Energy from observed structures statistically
  derived energies

7
Structure-derived potential KMB03

• Energy from observed structures as implemented
  in Rosetta

8
Three tasks in Hbond scoring

• Identify pairs of atoms that Hbond
• Classify Hbond types
• Evaluate energies for Hbonds
• Rosetta mixes these tasks together

9
Three tasks in Hbond scoring

• As described in KMB03
• Identify pairs of atoms that Hbond
• Params AH distance, ?, ?
• Classify Hbond types
• BB helix, strand, other AH distance
• SS,BS,SB acceptor hybridization AH dist
• Evaluate energies for Hbonds
• Sum three potentials on AH distance, ?, ?, ?
• Amino acid weights
• Residue neighbors for donor/acceptor

10
Three tasks in Hbond scoring

• As implemented in Rosetta
• Identify pairs of atoms that Hbond
• Params AH distance, ?, ?
• Classify Hbond types
• BB separation short sep4 long range
• SS,BS,SB acceptor hybridization AH dist
• Evaluate energies for Hbonds
• Sum three potentials on AH distance, ?, ?
• Amino acid weights OR Residue neighbors for
  donor/acceptor

11
SS bonds native decoy
dist
?
?
?
sp2ED QN bb
sp3TSY
ring H
12
SS bonds native decoy
dist
?
?
?

• Natives Dunbrack set of 3157 structures
• some pdb errors
• Decoys Best 20 for each of Rhijus ab initio
  runs on 62 structures
• small proteins
• few parallel beta strands
• Rosetta places Hs determines Hbonds
• Filter energies lt -0.1
• Visualization Tuftes small multiples
• Normalization
• Express counts as fraction of all Hbonds to
  support comparison of colors in each plot
• Plot with common x axis scale y to max height

sp2ED QN bb
sp3TSY
ring H
13
Energy distribution of bonds involving a
sidechain atom before/after filtering
14
Number (and percentage) of bonds under the
existing classification
Counts Counts Percentage Percentage
Native Decoys Native Decoys
BB Helix (/-4) 185,204 38,128 32.28 50.66
BB Turn (/-3) 79,110 8,983 13.79 11.94
BB Other 150,945 19,459 26.31 25.85
S sp2 ED QB bb 132,522 6,448 23.10 8.57
S sp3 TS Y 23,641 2,062 4.12 2.74
S ring H 2,325 184 0.41 0.24
TOTALS 573,747 75,264 100.00 100.00
15
Observations

• Rosetta does well at optimizing what it is told.
• Decoy distributions are more sharply peaked than
  natives.
• Relax preserves more non-helix bonds than ab
  initio, but produces same shapes for param
  distribns.
• To test changes, it suffices to run relax.

16
SS bonds native decoy
dist
?
?
?
sp2ED QN bb
sp3TSY
ring H
17
SS bonds native decoy
dist
?
?
?
sp2ED QN bb
sp3TSY
ring H
18
SS,BS,SB bonds native decoy
dist
?
?
?
sp2ED QN bb
sp3TSY
ring H
19

• AH Distance
• NB
• donor effects
• small
• omit C
• bimodal H
• R QNacc

20

• Theta A-H-D angle
• NB
• small s
• width
• R NENH

21

• PsiAHD angle
• NB
• RE EDacc

22

• Chi A2-A torsion
• NB
• Polar
• charged
• prefs

23

• __ H-D torsion
• NB
• Polar
• charged
• prefs

24
Three tasks in Hbond scoring

• Proposed changes
• Identify pairs of atoms that Hbond
• Params AH distance, ?, ?
• Classify Hbond types
• BB finer separation (Beta?)
• SS,BS,SB finer don/acc chemical types
• Evaluate energies for Hbonds options
• Sum three potentials on AH distance, ?, ?
• Potential on three variables AH distance, ?, ?
• Add neighbors
• Add a torsion as 4th or 5th variable
• Weights for tuning different terms

25
Backbone bonds
AH distance
26
Backbone bonds
theta
27
Parallel vs Anti-parallel beta

• The standard figures are misleading
• parallel and anti-parallel form similar distance,
  ?, ? distributions.

28
Backbone bonds
psi
29
Backbone bonds
chi torsion
30
Backbone bonds
AH-DD2 torsion
31
Refactoring Hbonds

• Recategorizing should eliminate long-range
  short-range Hbondswhich are used outside of
  hbonds.cc they shouldnt need to be.
• Duplicated code in minimizers needs to be brought
  back into hbonds.cc

32
Refactoring

• In code, a function should do one thing well.
• When a function you work with is doing too many
  things, split it.
• Duplicating code indicates that something is
  designed wrong.
• Avoid magic numbers.