Machine learning techniques can help address several unsolved problems in structural bioinformatics, including predicting protein flexibility and binding sites. The document discusses using machine learning models like SVMs trained on structural data to predict flexibility regions and protein-protein interaction sites from sequence alone. It also presents challenges in defining protein domain boundaries and predicting other structural features from sequence.

1. Machine Learning as Applied to Structural Bioinformatics: Results and Challenges Philip E. Bourne University of California San Diego [email_address]

2. The Current Situation Structure contributes greatly to our understanding of living systems We are locked into thinking about structure in specific ways which limits our view All too often we consider structure as a static entity The view at left is not how another protein or a small molecule ligand sees PKA We are still not very good at certain problems …

3. Example Unsolved Problems that Machine Learning Can Address Predicting flexibility and disorder in protein structure Predicting sites of protein-protein and protein-ligand interaction Predicting protein function Defining domain boundaries from sequence Predicting secondary, tertiary and quaternary structure Predicting what will crystallize

4. Example Unsolved Problems that Machine Learning Can Address Predicting flexibility and disorder in protein structure Predicting sites of protein-protein and protein-ligand interaction Predicting protein function Defining domain boundaries from sequence Predicting secondary, tertiary and quaternary structure Predicting what will crystallize * Will talk about this * Will offer as a challenge

5. The Current Situation: The Potential “Training Set” is Growing Quickly High level of redundancy as measured by sequence or structure Structure space is clearly very finite, but not clear how much is covered Increase in functionally uncharacterized structures Complexity is increasing, but still lack complexes Structures predominantly 1 and 2 domains Lack membrane proteins In summary the training set is still not truly representative but structural genomics will improve this situation

6. Predicting Functional Flexibility Jenny Gu Gu, Gribskov & Bourne PLoS Computational Biology 2006 Early On-line Release

7. Spectrum of Protein Order and Disorder Ordered Structures Disordered Structures If we believe that the 3-dimensional structure of a protein is defined by its 1-dimensional sequence then why not its flexibility?

8. Bridging the Sequence-flexibility Gap Generalize sequence - flexibility relationship to identify local protein regions important for allostery

9. The Training Dataset The dataset contains the following qualities: Non-redundant sequences training set with sequences containing ≤ 10% identity. With good quality structures R-factor < 0.30 At high resolution Resolution < 2.0 Å. Total number of proteins in dataset: 1277 sequences

10. Obtaining Protein Dynamic Information Protein structures treated as a 3-D elastic network. Bahar, I., A.R. Atilgan, and B. Erman Direct evaluation of thermal fluctuations in proteins using a single-parameter harmonic potential. Folding & Design, 1997. 2(3): p. 173-181.

11. Defining the Target Features Gaussian Network Model: Models protein structure as a 3-D elastic network. Each Ca is a node in the network. Each node undergoes Gaussian-distributed fluctuations influenced by neighboring interactions within a given cutoff distance. (7Å) Decompose protein fluctuation into a summation of different modes. Bahar, I., A.R. Atilgan, and B. Erman Direct evaluation of thermal fluctuations in proteins using a single-parameter harmonic potential. Folding & Design, 1997. 2(3): p. 173-181.

12. Side Note: Gaussian Network Model vs Molecular Dynamics GNM relatively cause grained GNM fast to compute vs MD Look over larger time scales Suitable for high throughput

13. Functional Flexibility Score Utilize correlated movements to help define regional flexibility with functional importance. Functionally Flexible Score For each residue: Find Maximum and Minimum Correlation Use to scale normalized fluctuation to determine functional importance

14. Example: Identifying Functional Flexible Regions (FFR) in HIV Protease Gu, Gribskov & Bourne PLoS Comp. Biol.. 2006 Early Release Correlated modes (yellow) Anti-correlated (blue) Normalized scores – single chain

15. Identifying Regions in Bovine Pancreatic Trypsin Inhibitor and Calmodulin

16. How to Represent the Protein Sequence? Residues characterized as FFs or not – approx 20% of residues with lengths typically 9+/-11 The longer the protein the longer the FFR We use hidden Markov models to represent each protein sequence in the training dataset. Hidden Markov models captures evolutionary information along with the probability of finding one of the 20 amino acids in each position of the sequence. Use probability states as input features in the first layer of an architecture containing two SVM layers.

17. Architecture of Wiggle Captures Evolutionary Effects Captures Local Effects (smoothing) 9*29 features used for each residue

18. Generating Additional Input Features Modified Bootstrapping – for Tripeptides – Accounts for Nearest Neighbors Effects Calculate Z score and P value for each pattern with respective null models Sample with replacement 44645 times Pooled Patterns (window size : 3) Null Model* for Non-FFR Regions Sample with replacement 199515 times * Generate 10,000 Null Models Null Model* for FFR Regions

19. Architecture of Wiggle Captures Evolutionary Effects Captures Local Effects (smoothing) 9*29 features used for each residue

20. Predictors Trained on the Entire Dataset Perform Poorly on Smaller Proteins. False Positive False Negative The characteristics of small proteins are different – eg percent of complexes

21. Partition Training Set Based on Sequence Length Prediction performance of SVM trained on a partitioned dataset (solid lines) is compared to that was trained on the entire dataset (dashed line). Prediction quality improved when dataset is partitioned. Most notably for proteins up to 200 amino acid residues long. Slight improvements observed for proteins longer than 200 residues. <200 AA Long >200 AA Long

22. Performance of Wiggle Predictors Wiggle Accuracy: 66.01% Precision: 37.11% Recall: 70.49% Wiggle 200 Accuracy: 76.46% Precision: 48.99% Recall: 78.27%

23. Case Study: PvuII Endonuclease FF SCORE (homodimer for DNA specific cleavage) Wiggle 200 Identify known loop for minor grove recognition Identify hinge residues not previously seen Important result for mutagenesis studies

24. Conclusions for Wiggle FFRs can be measured from structure With some empirical effort these data can be used as input to an SVM to predict FFRs from sequence alone Useful for: Improving docking studies Better understand protein function Engineer more or less stable proteins …… Gu, Gribskov & Bourne 2006 PLoS Comp. Biol.. 2006 Early Release

25. Exploiting Sequence and Structure Homologs to Identify Protein-Protein Binding Sites JoLan Chung Chung, Wang & Bourne 2006 Proteins: Structure, Function and Bioinformatics, 62(3) 630-640

26. Methods to Identify Protein-protein Binding Sites Docking Threading and homology modeling Evolutionary tracing Correlated mutations Properties of patches Hydrophobicity Neural networks and support vector machines (SVM)

27. None of the above methods consider the residues which are spatially conserved on the surfaces of structure homologs These residues are reported to correspond to the energy hot spots on protein interfaces and can be derived from multiple structure alignments Structurally Conserved Surface Residues?

28. Method: Incorporate Structural Conservation to Predict the Interface Residue Using SVM Support vector machine Sequence + structure information Binding site location

29. Derive the Structurally Conserved Residues The structural conservation scores were derived from multiple structural alignments and weighted by the normalized B-factors to consider the structure flexibility that will result in a bad alignment (could use FFRs in the future) Each position in the alignment has a structural conservation score, which represents the conservation in 3D space A position has a high conservation score if the aligned residues are spatially conserved

30. Structurally Conserved Residues and Interface Residues E.g. Residues with the top 20% of structure conservation scores (red) mapped to adrenodoxin (Adx, PDB code 1E6E:B) and known to bind adrenodoxin reductase (AR, blue).

31. Training D ataset 274 non-redundant chains of heterocomplexes (<30% sequence identity) extracted from the PDB Each of these chains was accompanied with a structure alignment with at least 4 members

32. SVM Training A surface residue ↓ Sequence profile + ASA + Structural conservation score in a window of 13 residues (The residue to be predicted and 12 spatially nearest surface residues) ↓ Support vector machine classifier ↓ Interface or non-interface residue ?

33. SVM Training Each residue was encoded as a feature vector with 13×21 dimensions: (the surface residue to be predicted + 12 nearest neighbors) x (20 amino acids + accessible surface area) Implemented using SVM light with the radial basis function as a kernel. (γ = 0.01, regularization parameter C =10) A set of non-interface surface residues was randomly selected to make the ratio of positive and negative data 1:1 3 fold cross-validation was performed

34. Predictor 1: Sequence profile + ASA. Predictor 2: Sequence profile + ASA + structural conservation score Predictor 3: Sequence profile + ASA + raw structural conservation score without weighted by the normalized B-factor Predictor 4: Sequence profile + ASA+ normalized B-factor The Performance of Various Predictors

35. Precise prediction: at least 70% interface residues were identified Correct prediction: at least 50 % interface residues were identified Partial prediction: some but less than 50 % interface residues were identified Wrong prediction: no interface residues were identified The Performances of the Predictors

36. Predicted Binding Sites - Example 1 Protein : domain 1 of the human coxsackie and adenovirus receptor (CAR D1) Mediate adenoviruses and coxsackie virus B infection CAR is an integral membrane protein expressed in a broad range of human and murine cell type. CAR D1 is one of its two extracellular domains Binding partner : knob domain of the adenoviruses serotype 12 (Ad12)

37. Predicted Binding Sites - Example 2 Protein : adrendoxin (Adx) In mitochondria of the adrenal cortex, the steroid hydroxylating system requires the transfer of electrons from the membrane-attached flavoprotein AR via the soluble Adx to the membrane-integrated cytochrome P450 of the CYP 11 family Binding partner : adrenodoxin reductase (AR)

38. Predicted Binding Sites - Example 3 Protein : fibroblast growth factor receptor 2 (FGFR2) Ser252Trp Mutant Apert syndrome (AS) is caused by substitution of one of two adjacent residues, Ser252Trp or Pro253Arg Binding partner : fibroblast growth factor (FGF2)

39. Conclusions – Protein-protein Binding Sites Incorporating the structural conservation score improved the prediction performance of SVM significantly This study is an initial trial that exploits multiple structure alignment for the large scale prediction of functional regions We need better algorithms for multiple structure alignment (we have one benchmark for anyone interested) This method can be used to guide experiments, such as site-specific mutagenesis, or combined with docking procedures to limit the search space

40. General Conclusions Using known features of protein structure these can be mapped to the corresponding sequences and used to train an SVM Having evaluated the SVM in a cross validation tests the performance can be determined Good performance is shown in training for both flexibility and sites of protein-protein interaction These predictors are currently being used to solve real biological problems Can this approach be applied to other aspects of structure?

41. 1fohb PUU: 2 Experts: 3 A. B. C. D. E. Consider Domain Definitions: Holland et al. 2006 JMB Early Release Veretnik et al. 2004 JMB 339(3), 647-678 1ytf PUU: 1 Experts: 2 1d0gt PUU: 1 Experts: 3 1dgk PUU: 6 Experts: 4 1aoga PUU: 4 Experts: 3

42. Challenge – Defining Domain Boundaries from Sequence A domain is the unit of currency of proteins – domain structures define function, indicate evolutionary relationships etc… Domain prediction from structure easier than from sequence, but still not a solved problem Recently developed an accurate test set of domain definitions and boundaries: http://pdomains.sdsc.edu Good luck! Benchmark Data Available See: Holland et al 2006 JMB Early Release

43. Acknowledgements Functional Flexibility Jenny Gu & Michael Gribskov Protein-protein Interactions JoLan Chung & Wei Wang Domain Definitions Stella Veretnik, Tim Holland, Ilya Shindalov, Nick Alexandrov, Abdur Sikur Funding, NSF, NIH

44. The structural conservation score Raw structural conservation score where if a is not gap and b is not gap otherwise where N is the total number of aligned structures, s i ( x ) is the amino acid at position x in the i th structure in the alignment, m is a modified PET substitution matrix calculated by Valdar et al.

45. The structure conservation score The B-factors determined by X-ray crystallographic experiments provide an indication of the degree of mobility and disorder of an atom in a protein structure Raw structural conservation scores were weighted by the normalized B-factors ( B norm, i ) to consider the structure flexibility where