Volume 10 Supplement 1

## Selected papers from the Seventh Asia-Pacific Bioinformatics Conference (APBC 2009)

- Research
- Open Access

# Finding motif pairs in the interactions between heterogeneous proteins via bootstrapping and boosting

- Jisu Kim
^{1}, - De-Shuang Huang
^{2}and - Kyungsook Han
^{1}Email author

**10 (Suppl 1)**:S57

https://doi.org/10.1186/1471-2105-10-S1-S57

© Kim and Han; licensee BioMed Central Ltd. 2009

**Published:**30 January 2009

## Abstract

### Background

Supervised learning and many stochastic methods for predicting protein-protein interactions require both negative and positive interactions in the training data set. Unlike positive interactions, negative interactions cannot be readily obtained from interaction data, so these must be generated. In protein-protein interactions and other molecular interactions as well, taking all non-positive interactions as negative interactions produces too many negative interactions for the positive interactions. Random selection from non-positive interactions is unsuitable, since the selected data may not reflect the original distribution of data.

### Results

We developed a bootstrapping algorithm for generating a negative data set of arbitrary size from protein-protein interaction data. We also developed an efficient boosting algorithm for finding interacting motif pairs in human and virus proteins. The boosting algorithm showed the best performance (84.4% sensitivity and 75.9% specificity) with balanced positive and negative data sets. The boosting algorithm was also used to find potential motif pairs in complexes of human and virus proteins, for which structural data was not used to train the algorithm. Interacting motif pairs common to multiple folds of structural data for the complexes were proven to be statistically significant. The data set for interactions between human and virus proteins was extracted from BOND and is available at http://virus.hpid.org/interactions.aspx. The complexes of human and virus proteins were extracted from PDB and their identifiers are available at http://virus.hpid.org/PDB_IDs.html.

### Conclusion

When the positive and negative training data sets are unbalanced, the result via the prediction model tends to be biased. Bootstrapping is effective for generating a negative data set, for which the size and distribution are easily controlled. Our boosting algorithm could efficiently predict interacting motif pairs from protein interaction and sequence data, which was trained with the balanced data sets generated via the bootstrapping method.

## Keywords

- Feature Vector
- Virus Protein
- Linear Motif
- Residue Pair
- Negative Data

## Background

Linear motifs are known to facilitate many protein-protein interactions [1]. Despite the availability of a large volume of data about protein-protein interactions and their sequences, linear motifs are difficult to discover, due to their short length, which is between three and ten amino acids [2]. Recently, several methods have been developed for discovering linear motifs of protein-protein interactions [1, 3], but most methods focus on detecting individual linear motifs rather than interacting motif pairs. Motif pairs are more useful than motifs for filtering many spurious protein interactions in current high-throughput data, and for identifying a functional target.

Supervised learning or stochastic methods are often used to predict linear motifs involved in protein-protein interactions. Both negative and positive interactions are required to train the methods. Unlike positive interaction data, negative samples cannot be readily obtained from protein-protein interaction data. Assuming a negative interaction where there is no explicit evidence of a positive interaction results in a much larger negative data set than a positive data set. Such an unbalance between positive and negative data sets makes a prediction biased [4, 5]. Generating a negative data set via random selection often does not reflect the original distribution of data, thus it does not produce a good prediction model.

There are a few methods for generating a negative data set. Jansen et al. [6] generate a data set of negative interactions by assuming that proteins in different subcellular compartments of a cell do not interact. However, different subcellular locations only indicate that the proteins have a lower chance of binding than those in the same location, and some proteins are found in more than one subcellular compartment of a cell [7]. The method developed by Gomez et al. [8] assumes a negative protein interaction, if there is no explicit evidence of an interaction. However, this assumption generates a negative data set that is too large, resulting in low sensitivity in interaction predictions. The method that uses the shortest path [7] has difficulty in obtaining a negative data set of the desired size. The method that uses sequence similarity [9] also has difficulty in controlling the size of the negative data set.

In this study, we developed a bootstrapping algorithm for generating a negative data set of protein-protein interactions, and a new boosting algorithm for finding interacting motif pairs from positive and negative data sets. The remainder of the paper describes the algorithms and their experimental results with various parameter values.

## Results and discussion

In the following description, the *sampling size S* is the number of negative samples that were examined to generate a single negative data via bootstrapping. When the number of negative samples with *m*-th feature = 1 is greater than the *acceptance ratio A*, the *m*-th feature of the re-sampled negative data is set to 1. The feature vector and the acceptance ratio are described in detail in the method section.

### Affect of acceptance ratios

_{i}= {motif pairs found in at least

*i*folds} where i = {1, 2, ..., 5} [7]. Table 1 shows the number of motif pairs predicted with different acceptance ratios.

Motif pairs found during five-fold cross validation

A = 1/10 | A = 1/8 | A = 1/6 | A = 1/4 | |
---|---|---|---|---|

M | 12563 | 21821 | 50634 | 142395 |

M | 3479 | 4866 | 12472 | 38008 |

M | 1047 | 1181 | 3498 | 15220 |

M | 189 | 344 | 874 | 6970 |

M | 28 | 105 | 141 | 2134 |

As the acceptance ratio increases, re-sampled negative data have fewer nonzero features, resulting in more motif pairs. This is because the nonzero features of negative data are used to filter out the features that are also nonzero in positive data.

Prediction performance with respect to acceptance ratios of bootstrapping

A = 1/10 | A = 1/8 | A = 1/6 | A = 1/4 | |
---|---|---|---|---|

Sensitivity | 58.35% | 75.88% | 82.42% | 90.42% |

Specificity | 78.83% | 84.40% | 92.29% | 96.02% |

Accuracy | 66.09% | 80.14% | 87.35% | 93.22% |

### Affect of proportions of positive and negative data sets

Prediction performance with respect to proportions of positive and negative data

Data ratio (P: N) | 1712: 2283 (2: 3) | 1712: 1712 (1: 1) | 1712: 1141 (3: 2) |
---|---|---|---|

Sensitivity | 68.98% | 75.88% | 77.80% |

Specificity | 87.03% | 84.40% | 77.56% |

Accuracy | 79.30% | 80.14% | 77.70% |

### Affect of boosting algorithms

Prediction performance of two boosting algorithms

Boosting algorithm | AdaBoost algorithm | Our Boosting algorithm |
---|---|---|

Sensitivity | 70.55% | 75.88% |

Specificity | 84.21% | 84.40% |

Accuracy | 77.37% | 80.14% |

### Motif pairs found in complexes of human and virus proteins

_{1}= 1, implying that motif pairs of M

_{1}had no more significance than random motif pairs. However, motif pairs of M

_{2}-M

_{5}were more significant than random motif pairs. Figure 2 shows a complex of human and HIV-1 proteins (PDB ID: 1AGF). Among the total of 63 contact residues between chains A and C, 16 residue pairs were included in M

_{2}.

Motif pairs found in each fold

Set | # of motif pairs | p-value |
---|---|---|

M | 334 | 1 |

M | 87 | 3.13e-3 |

M | 22 | 3.02e-3 |

M | 7 | 2.25e-2 |

M | 2 | 1.79e-1 |

## Conclusion

When positive and negative training data sets are unbalanced, the result via the prediction model tends to be biased. We developed a bootstrapping algorithm for generating a negative data set of arbitrary size from protein-protein interaction data. We also developed an efficient boosting algorithm for finding interacting motif pairs in human and virus proteins. The boosting algorithm showed the best performance (84.4% sensitivity and 75.9% specificity) with balanced positive and negative data sets. The boosting algorithm was also used to find potential motif pairs in complexes of human and virus proteins, for which structural data was not used for training the algorithm. Interacting motif pairs common to multiple folds of structural data of complexes were proven to be statistically significant.

This method predicts protein-protein interactions and motif pairs using the protein sequence data. The sequence information alone is insufficient to predict motif pairs for some proteins, but our method provides a useful model for predicting motif pairs in protein-protein interactions when the sequence is the only information available. The data set for interactions between human and virus proteins was extracted from BOND and is available at http://virus.hpid.org/interactions.aspx. The complexes of human and virus proteins were extracted from PDB and their identifiers are available at http://virus.hpid.org/PDB_IDs.html.

## Methods

### Data set

We extracted the latest data of interactions between human and virus proteins from BOND [10]. As of May, 2008, there were 1,712 interactions between 1,029 human proteins and 603 virus proteins. These interactions were considered as positive data. From 1,712 interactions, we constructed three negative data sets of 2,252, 1,712, and 2,283 samples via the bootstrapping method.

### Feature vector

The way of extracting features in our study was similar to the one used in the studies of Gomez et al. [8] and Yu et al. [7]. In the study by Gomez et al., four-tuple features were used to identify a subsequence of four amino acids. Based on biochemical similarities of amino acids, twenty amino acids were classified into six categories: {IVLM}, {FYW}, {HKR}, {DE}, {QNTP}, and {ACGS} [11]. After classification, there were 6^{4} = 1,296 possible substrings of length four.

Encoding scheme for the interacting motif pairs

Biochemical property | 4-tuple pairs (M bits) | |||
---|---|---|---|---|

Candidate motif pair | ||||

Classification | Category number | Bit number | Human 4-tuple | Virus 4-tuple |

{I, V, L, M} | 0 | 1 | 0000 | 0000 |

{F, Y, W} | 1 | 2 | 0000 | 0001 |

{H, K, R} | 2 | ⋮ | ⋮ | ⋮ |

{D, E} | 3 | |||

{Q, N, T, P} | 4 | M-1 | 5555 | 5554 |

{A, C, G, S} | 5 | M | 5555 | 5555 |

_{symmetric}of possible motif pairs is determined by

The interactions between human and virus proteins are the interactions between heterogeneous proteins. Hence, the total number M_{asymmetric} of possible motif pairs is as follows.

*M*_{
asymmetric
}= 6^{4}·6^{4} = 1,679,616

Our method is intended for finding motif pairs with 4 consecutive residues (i, i+1, i+2 and i+3) in each motif. Hence, a motif with non-consecutive residues cannot be found even if the residues are spatially close to each other. Since the total number of possible motif pairs is 6^{
m
}·6^{
m
}= (6^{
m
})^{2} = 6^{2m}for a motif of size *m* (equation 5), the total number of possible motif pairs increases exponentially as the size of *m* increases. The total number of possible motif pairs can be reduced with a motif of a smaller size (e.g., 2 or 3 residues), but the motif of a small size has too many occurrences in the sequences, which significantly reduces the selectivity of the motif.

### Bootstrapping for re-sampling

As in Gomez et al.'s method [8], we assumed a negative interaction if there was no explicit evidence of an interaction. However, this assumption generates a much larger number of negative samples than positive samples. If we randomly select only some of the negative samples, we might miss information from unselected negative samples. Dupret and Koda [5] used bootstrapping to identify the optimal re-sampling proportions in binary classification experiments.

In our study, we used bootstrapping to generate negative data sets via re-sampling negative data. Algorithm 1 describes our bootstrapping method, which is controlled by the sampling size *S* and acceptance ratio *A*. Executing the bootstrapping algorithm yields a single re-sampled negative data from *S* negative data. The re-sampled negative data is represented as a feature vector Y = {y_{1}, y_{2}, ..., y_{M}} via Algorithm 1. The number of 1's in the feature vector Y is controlled by the acceptance ratio *A*. A larger value of *A* produces a feature vector with fewer nonzero elements.

### Algorithm 1 – Bootstrapping algorithm

This algorithm generates the feature vector Y for a single negative data from S samples, where S is the sampling size and A is the acceptance ratio for setting a feature to 1.

1. Randomly sample S protein pairs (P_{s1}, P_{s2}) with replacement from non-interacting protein pairs, where s = {1, 2, ..., S}.

2. Initialize n_{i} = 0 for i = {1, 2, ..., M}

3. Initialize y_{i} = 0 for i = {1, 2, ..., M}

4. For s == {1, 2, ..., S}

a. Make a binary vector X_{s} = {x_{s1}, x_{s2}, ..., x_{sM}} for a pair of proteins (P_{s1}, P_{s2})

b. For m = {1...M}

If x_{sm} = 1, n_{m} = n_{m} + 1 {n_{m} is the number of samples for which the *m*-th feature = 1}

5. For m = {1...M}

If n_{m}/S > A, set y_{m} = 1

6. Y = {y_{1}, y_{2}, ..., y_{M}} is a feature vector representing re-sampled negative data.

### The boosting algorithm

In general, the boosting method finds a highly accurate hypothesis by combining weak hypotheses, each of which is only moderately accurate. Typically, each weak hypothesis is a simple classification rule. In AdaBoost (Adaptive Boosting), each weak hypothesis generates not only a classification rule but also a confidence score that estimates the reliability of the classification [12].

The study of Yu et al. [7] uses the AdaBoost algorithm for finding motif pairs in homogeneous protein interactions. One of the differences between Yu's algorithm and ours is the number of weak hypotheses used in the algorithms. In Yu's AdaBoost algorithm, if the weight (α_{s1}) of the first weak hypothesis is much greater than the weights of other hypotheses, the final hypothesis is determined mainly by the first weak hypothesis and other hypotheses have negligible effect on the final hypothesis.

Our boosting algorithm determines the weights of weak hypotheses and uses the training data in a different way from Yu's algorithm. While Yu's AdaBoost algorithm uses different weights and the same training data per weak hypothesis, our algorithm uses the same weights and different training data per weak hypothesis. Our boosting algorithm uses fewer weak hypotheses than Yu's algorithm, and requires much less time than their algorithm.

Our algorithm consists of two parts: boosting algorithm and WINNOW2 algorithm. The boosting algorithm described in Algorithm 2 takes as input a training set (x_{1}, y_{1}), ..., (x_{n}, y_{n}), where each pair is a binary vector of length M, which represents an interaction with a label in the label set Y. Y = {-1, +1} indicates whether each interaction is positive or negative. The boosting algorithm calls the WINNOW2 algorithm to obtain a weak hypothesis in an iterative series of rounds, where t = {1, ..., S}. In each round, the boosting algorithm computes the weight (α_{t}) of the weak hypothesis *h*_{c,t}. The final hypothesis *H*_{
t
}for *Set*_{
t
}is the weighted sum of weak hypotheses *h*_{c,i}(*i* = 1, ..., *S* and *i* ≠ *t*).

We used a regulated stochastic WINNOW2 algorithm [13] with R = 200,000 as a weak classifier (Algorithm 3). The WINNOW2 algorithm is similar to that of Yu et al. [7], except for the step of updating learner factors. Yu's algorithm updates learner factors when x_{ki} (feature vector) is 0, but our algorithm updates them when x_{ki} is 1. Yu's algorithm takes as input a training set and computes normalized sample weights in each boosting round. In the step of drawing a sample data, data with larger weights are drawn more frequently than those with smaller weights. Since the sample weights are difficult to adjust in each round, our algorithm uses the same weight for every sample and draws samples with equal frequency. But, the training data is changed in every round, and the call to the WINNOW2 algorithm produces different hypotheses according to the training data. Finally, additional regulation is performed to discover effective components. The components with large learner factors are identified as effective components. These effective components are considered as the motif pairs of protein-protein interactions.

### Algorithm 2 – boosting algorithm

The boosting algorithm calls the WINNOW2 algorithm to obtain weak hypotheses. S is the number of divided data sets.

1. Given divided data set *Set*_{1}, *Set*_{2}, ..., *Set*_{
S
}where $\underset{t=1}{\overset{S}{\cup}}Se{t}_{t}}=Se{t}_{total$.

2. For *t* = 1, ..., *S*

a. Given training data (*x*_{1}, *y*_{1}), (*x*_{2}, *y*_{2}), ..., (*x*_{
n
}, *y*_{
n
}) from *Set*_{
t
}where *x*_{
i
}∈ {0, 1}^{
M
}, *y*_{
i
}∈ *Y* = {-1, +1} for {*i* = 1, 2, ..., *n*}

b. Call the WINNOW2 algorithm to obtain the weak hypothesis *h*_{c,t}.

*r*

_{ t }of the weak hypothesis

*h*

_{c,t}at level c.

*α*

_{ t }of the weak hypothesis

*Set*

_{ t }:

### Algorithm 3 – WINNOW2 algorithm

The WINNOW2 algorithm trains the weak hypothesis. R is the number of randomly selected data.

1. Given training data (*x*_{1}, *y*_{1}), (*x*_{2}, *y*_{2})..., (*x*_{
n
}, y_{
n
}).

2. Initialize learner factor *w*_{
i
}= 1 for *i* = {1, 2, ..., *M*}, and threshold *θ* = *M*/2

3. For *r* = {1, ..., *R*}

a. Randomly select a sample data (*x*_{
k
}, *y*_{
k
}), and let vector *x*_{
k
}denote (*x*_{k 1}, *x*_{k 2}, ..., *x*_{
kM
})

c. Update learner factors ${w}_{i}={w}_{i}{2}^{{x}_{ki}(y-h)/2}$

*h*

_{ c }at level

*c*as follows:

where *w*_{i,c}= *w*_{
i
}if *w*_{
i
}≥ *c*, and *w*_{i,c}= 0 otherwise.

5. Let *N*_{
c
}denote the number of positive predictions by classifier *h*(*c*) in the training data and *N*_{0} denote the number of positive predictions with the cutoff of 0.

Output the classifier *h*_{
C
}where *C* = arg max {*c* | *N*_{
c
}= *N*_{0}}.

6. The features with non-zero w_{i,c} are effective motif pairs.

## Verification with structural data

To further evaluate the algorithm for the structures of heterogeneous multi-protein complexes, we extracted structural data for complexes of human and virus proteins from PDB [14]. Complexes with RNA or DNA chains were not retrieved. Circa June 2008, there were a total of 105 complexes of human and virus proteins in PDB.

We used five-fold cross validation to evaluate the algorithm. The data set was split into five parts of equal size. The boosting algorithm using the WINNOW2 algorithm for weak hypotheses was trained with one part and tested with the remaining four parts. The train-test procedure consisted of five iterations.

When a residue pair in different chains contained an atomic pair within the distance of 5 Å, we considered the residue pair as a *contact residue pair*. If a motif pair had at least one contact residue pair, we considered the motif pair as a *verifiable motif pair* [7]. To assess the statistical significance of motif pairs predicted by our algorithm, we estimated the p-value of motif pairs by executing Algorithm 4 with *m* = 100,000 [9]. Motif pairs with lower p-values are more significant than those with higher p-values.

### Algorithm 4 – Estimation of p-values of motif pairs

A motif pair with a smaller p-value is more significant than a random motif pair R_{i}.

1. Given a set *S* of motif pairs collected by weak hypotheses.

2. Randomly draw *m* motif pairs {*R*_{1}, *R*_{2}, ..., *R*_{
m
}} where *R*_{
i
}has the same size as M_{k} (k = 1, 2, ...., 5)

where *V*(*S*) is the number of verifiable motif pairs.

## Declarations

### Acknowledgements

This work was supported by the Korea Research Foundation Grant funded by the Korean Government (KRF-2006-D00038).

This article has been published as part of *BMC Bioinformatics* Volume 10 Supplement 1, 2009: Proceedings of The Seventh Asia Pacific Bioinformatics Conference (APBC) 2009. The full contents of the supplement are available online at http://www.biomedcentral.com/1471-2105/10?issue=S1

## Authors’ Affiliations

## References

- Davey NE, Shields DC, Edwards RJ: SLiMDisc: short, linear motif discovery, correcting for common evolutionary. Nucleic Acid Res. 2006, 34: 3546-3554. 10.1093/nar/gkl486.PubMed CentralView ArticlePubMedGoogle Scholar
- Neduva V, Russel RB: Linear motifs: Evolutionary interaction switches. FEBS Letters. 2005, 579: 3342-3345. 10.1016/j.febslet.2005.04.005.View ArticlePubMedGoogle Scholar
- Neduva V, Russel RB: DILIMOT: discovery of linear motifs in proteins. Nucleic Acid Res. 2006, 34 (Web Server issue): W350-W355. 10.1093/nar/gkl159.PubMed CentralView ArticlePubMedGoogle Scholar
- Olson DL: Data Set Balancing. Lecture Notes in Artificial Intelligence. 2004, 3327: 71-80.Google Scholar
- Dupret G, Koda M: Bootstrap re-sampling for unbalanced data in supervised learning. European Journal of Operational Research. 2001, 134: 141-156. 10.1016/S0377-2217(00)00244-7.View ArticleGoogle Scholar
- Jansen R, Gerstein M: Analyzing protein function on a genomic scale: the importance gold-standard positives and negatives for network prediction. Current opinion in Microbiology. 2004, 7: 535-545. 10.1016/j.mib.2004.08.012.View ArticlePubMedGoogle Scholar
- Yu H, Qian M, Deng M: Using a Stochastic AdaBoost Algorithm to Discover Interactome Motif Pairs from Sequences. Lecture Notes in Bioinformatics. 2006, 4115: 622-630.Google Scholar
- Gomez SM, Noble WS, Rzhetsky A: Learning to Predict Protein-Protein Interactions from Protein Sequences. Bioinformatics. 2003, 19: 1875-1881. 10.1093/bioinformatics/btg352.View ArticlePubMedGoogle Scholar
- Kim J, Park B, Han K: Prediction of Interacting Motif Pairs using Stochastic Boosting. Proceedings of Frontiers in the Convergence of Bioscience and Information Technologies. 2007, 95-100.Google Scholar
- Alfarano C, Andrade CE, Anthony K: The Biomolecular Interaction Network Database and related tools 2005 update. Nucleic Acid Res. 2005, 33 (Database issue): D418-D424. 10.1093/nar/gki051.PubMed CentralView ArticlePubMedGoogle Scholar
- Taylor WR, Jones DT: Deriving an amino acid distance matrix. Journal of Theoretical Biology. 1993, 164: 65-83. 10.1006/jtbi.1993.1140.View ArticlePubMedGoogle Scholar
- Schapire RE, Singer Y: Improved Boosting Algorithms Using Confidence-rated Predictions. Machine Learning. 1999, 37: 297-336. 10.1023/A:1007614523901.View ArticleGoogle Scholar
- Littlestone N: Learning Quickly When Irrelevant Attributes Abound. A New Linear-threshold Algorithm. Machine Learning. 1988, 2: 285-318.Google Scholar
- Deshpande N, Addess KJ, Bluhm WF: The RCSB Protein Data Bank: a redesigned query system and relational database based on the mmCIF schema. Nucleic Acids Research. 2005, 33: D233-D237. 10.1093/nar/gki057.PubMed CentralView ArticlePubMedGoogle Scholar

## Copyright

This article is published under license to BioMed Central Ltd. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.