Prediction of protein domain with mRMR feature selection and analysis.

Abstract

The domains are the structural and functional units of proteins. With the avalanche of protein sequences generated in the postgenomic age, it is highly desired to develop effective methods for predicting the protein domains according to the sequences information alone, so as to facilitate the structure prediction of proteins and speed up their functional annotation. However, although many efforts have been made in this regard, prediction of protein domains from the sequence information still remains a challenging and elusive problem. Here, a new method was developed by combing the techniques of RF (random forest), mRMR (maximum relevance minimum redundancy), and IFS (incremental feature selection), as well as by incorporating the features of physicochemical and biochemical properties, sequence conservation, residual disorder, secondary structure, and solvent accessibility. The overall success rate achieved by the new method on an independent dataset was around 73%, which was about 28-40% higher than those by the existing method on the same benchmark dataset. Furthermore, it was revealed by an in-depth analysis that the features of evolution, codon diversity, electrostatic charge, and disorder played more important roles than the others in predicting protein domains, quite consistent with experimental observations. It is anticipated that the new method may become a high-throughput tool in annotating protein domains, or may, at the very least, play a complementary role to the existing domain prediction methods, and that the findings about the key features with high impacts to the domain prediction might provide useful insights or clues for further experimental investigations in this area. Finally, it has not escaped our notice that the current approach can also be utilized to study protein signal peptides, B-cell epitopes, HIV protease cleavage sites, among many other important topics in protein science and biomedicine.

Full text

Page 1

Prediction of Protein Domain with mRMR Feature
Selection and Analysis
Bi-Qing Li2,3., Le-Le Hu1., Lei Chen4, Kai-Yan Feng3, Yu-Dong Cai1,5*, Kuo-Chen Chou5*
1Institute of Systems Biology, Shanghai University, Shanghai, China, 2Key Laboratory of Systems Biology, Shanghai Institutes for Biological Sciences, Chinese Academy of
Sciences, Shanghai, China, 3Shanghai Center for Bioinformation Technology, Shanghai, China, 4College of Information Engineering, Shanghai Maritime University,
Shanghai, China, 5Gordon Life Science Institute, San Diego, California, United States of America
Abstract
The domains are the structural and functional units of proteins. With the avalanche of protein sequences generated in the
postgenomic age, it is highly desired to develop effective methods for predicting the protein domains according to the
sequences information alone, so as to facilitate the structure prediction of proteins and speed up their functional
annotation. However, although many efforts have been made in this regard, prediction of protein domains from the
sequence information still remains a challenging and elusive problem. Here, a new method was developed by combing the
techniques of RF (random forest), mRMR (maximum relevance minimum redundancy), and IFS (incremental feature
selection), as well as by incorporating the features of physicochemical and biochemical properties, sequence conservation,
residual disorder, secondary structure, and solvent accessibility. The overall success rate achieved by the new method on an
independent dataset was around 73%, which was about 28–40% higher than those by the existing method on the same
benchmark dataset. Furthermore, it was revealed by an in-depth analysis that the features of evolution, codon diversity,
electrostatic charge, and disorder played more important roles than the others in predicting protein domains, quite
consistent with experimental observations. It is anticipated that the new method may become a high-throughput tool in
annotating protein domains, or may, at the very least, play a complementary role to the existing domain prediction
methods, and that the findings about the key features with high impacts to the domain prediction might provide useful
insights or clues for further experimental investigations in this area. Finally, it has not escaped our notice that the current
approach can also be utilized to study protein signal peptides, B-cell epitopes, HIV protease cleavage sites, among many
other important topics in protein science and biomedicine.
Citation: Li B-Q, Hu L-L, Chen L, Feng K-Y, Cai Y-D, et al. (2012) Prediction of Protein Domain with mRMR Feature Selection and Analysis. PLoS ONE 7(6): e39308.
doi:10.1371/journal.pone.0039308
Editor: Bin Xue, Uni. of South Florida, United States of America
Received November 27, 2011; Accepted May 17, 2012; Published June 15, 2012
Copyright: ? 2012 Li et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted
use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. This work was supported by
grants from the National Basic Research Program of China (2011CB510102, 2011CB510101) and the Innovation Program of Shanghai Municipal Education
Commission (12ZZ087).
Competing Interests: The authors have declared that no competing interests exist.
* E-mail: cai_yud@yahoo.com.cn (YDC); kcchou@gordonlifescience.org (KCC)
. These authors contributed equally to this work.
Introduction
Protein domains are structural, evolutionary and functional
units of proteins. Prediction of protein domains from the sequence
information can facilitate the prediction of protein tertiary
structure [1,2], the annotation of protein functions [2,3], the
protein structure determination [4], protein engineering [5] as well
as mutagenesis [6,7]. Particularly, the functional domains are
actually the cores of proteins that play the major role for their
functions. That is why in determining the 3D (three dimensional)
structure of a protein by experiments (see, e.g., [8,9,10,11]) or by
computational modeling (see, e.g., [7,12,13,14,15]) the first
priority was always focused on its functional domain. This is
because the knowledge of protein functional domains is important
for both basic research and drug development. Recently, the
functional domain information of proteins has been widely used to
formulate protein samples through the conception of pseudo
amino acid composition [16,17] for predicting various important
attributes of proteins, such as membrane proteins and their types
[18], GPCRs and their types [19,20], proteases and their types
[21], protein quaternary structural attribute [22,23], protein
structural classification [24], and protein subcellular localization
[25,26,27]. Meanwhile, the protein domain information was also
used to help analyzing protein-protein binding interactions [28,29]
and predicting the network of substrate-enzyme-product triads
[30].
With the avalanche of protein sequences generated in the
postgenomic age, many efforts have been made in hopes to predict
the domains of proteins from their primary sequences alone. They
can be roughly divided into three categories: (i) template-based
method [31,32,33], (ii) ab-initio method [34,35], and (iii) hybrid
method by combining the aforementioned two [36,37,38]. Most
template-based approaches attempted to find homologous se-
quences in the existing domain databases and then infer the
domains of the query protein from these sequences. The obvious
drawback of the template-based method was that it would work
only when a domain was conserved and had already been
deposited in a database. In other words, such an approach would
fail to work if the query protein did not have significant sequence
similarity to any of the domain-known proteins. In contrast to the
PLoS ONE | www.plosone.org1 June 2012 | Volume 7 | Issue 6 | e39308

Page 2

template-based method, the ab-initio method could make predic-
tions basically only based on the primary sequence information
alone, and hence for those query proteins without significant
sequence similarity to any of the domain-known proteins, the ab-
initio method would be a good choice. The concreted techniques
involved in the ab-initio methods are the machine learning
algorithms [35,39], artificial neural networks [40], and support
vector machines [41,42], along with the high quality domain
databases such as CATH [43], SCOP [44] and DALI [45].
However, since it needed to scan the entire sequence of a protein
usually involving several hundreds of amino acids, and also relied
on the inputs containing weak domain information, the ab-initio
method needed much more computational time and also often
suffered from low prediction accuracy. The third method, or the
hybrid method [36,37,46], took the strategy by integrating the
template-based method and the ab-initio method. In the mean
time, many softwares and web-server tools were developed for
predicting protein domains, such as FIEFDom [47], DoMpro
[40], DROP [42], DomCut [48] and Globplot [49]. Most of these
tools aimed at predicting domain linker or domain boundary, and
then followed by inferring the domain region.
The present study was initiated in an attempt to address the
problem from such a keypoint by first identifying each of the
constituent amino acid residue in a query protein belonging to the
domain region or not. The techniques involved were RF (random
forest), mRMR (maximum relevance minimum redundancy), and
IFS (incremental feature selection). The amino acid features
incorporated were the sequence conservation, residual disorder,
secondary structure propensity, and solvent accessibility.
As summarized in a review [17] and demonstrated by a series of
recent publications [50,51,52,53,54,55,56], to establish a really
useful statistical predictor for a protein system, we need to consider
the following procedures: (i) construct or select a valid benchmark
dataset to train and test the predictor; (ii) formulate the protein
samples with an effective mathematical expression that can truly
reflect their intrinsic correlation with the target to be predicted; (iii)
introduce or develop a powerful algorithm (or engine) to operate
the prediction; (iv) properly perform cross-validation tests to
objectively evaluate the anticipated accuracy of the predictor.
Below, let us describe how to deal with these steps.
Materials and Methods
1. Benchmark Dataset
A total of 517,100 protein sequences were retrieved from
UniProt/Swiss-Prot database (version 2010_06) [57]. In order to
construct a high-quality benchmark dataset, protein sequences
were screened strictly according to the following criteria. (i) To
reduce redundancy and homology bias, the cutoff threshold was
set at 25% as suggested in [58], meaning that those sequence
samples were removed by means of the program CD-HIT [59]
that had §25% pairwise sequence identity to any other in the
dataset. (ii) Of the remaining 45,942 protein samples obtained via
the above winnowing procedure, only 9,409 were kept that had
clear experimental domain annotations. (iii) Of the samples
obtained via the above step, 110 proteins were removed because
their disorder feature could not be calculated. Finally, a total of
9,299 protein sequences were obtained for the benchmark dataset
S used in this study.
Furthermore, on the basis of the benchmark dataset S, two
working datasets, i.e., a learning (training) dataset SLand an
independent testing dataset ST, were constructed. In order to fully
use the data in S and meanwhile guarantee that SLand STbe
completely independent of each other, the following condition was
imposed:
SL|ST~S and SL\ST~1
ð1Þ
where |, \, and 1 represent the symbols for ‘‘union’’,
‘‘intersection’’, and ‘‘empty set’’ in the set theory, respectively.
Constrained by the condition of Eq.1, 8,000 protein sequences
were randomly picked for the learning dataset SLand the
remaining 1,299 sequences for the testing dataset ST. See the
Online Supporting Information S1 for the codes of the proteins
included in the two datasets, SLand ST, respectively.
Three different sliding windows [60] were used to generate the
positive and negative datasets for this study: size-13, size-15, and
size-17. For the size-13 window, we extracted all the 13-residue
segments along a protein chain. The segments thus obtained can
be denoted as seg(13) and classified into the following two groups:
seq(13)~
positive,
if the center residue at the subsite
7 is within the domain region
otherwisenegaticve,
8
:
<
ð2Þ
During the operation of sliding the window along a protein
chain (cf. Figure 4 of [61], not all segments thus generated contain
13 amino acid residues. For those with less than 13 residues such
as the ones generated at the positions close to the N-terminal or C-
terminal, we complement their subsites with the nominal amino
acid ‘‘X’’ to make them contain 13 residues as well. Thus, we
obtained 1,694,782 positive samples and 4,093,531 negative
samples from the learning dataset SL. Subsequently, for each of
the two sets of 13-residue samples, the program CD-HIT [59] was
used to remove those that had §40% pairwise sequence identity
to any other in a same set. Finally, we obtained 121,013 positive
samples and 242,026 negative samples; i.e.,
Sz
S{
13contains 121,013 positive segments of seq 13
13contains 242,206 negative segments of seq 13
ðÞ
ðÞ
(
ð3Þ
where Sz
using the size-13 sliding window according to Eq.2, while S{
corresponding negative dataset derived from SL.
By following the same procedure but using size-15 and size-17
sliding windows, respectively, we obtained
13represents the positive learning dataset derived from SL
13the
Sz
S{
15contains 88,056 positive segments of seq(15)
15contains 176,112 negative segments of seq(15)
(
ð4Þ
and
Sz
S{
17contains 89,044 positive segments of seq(17)
17contains 178,088 negative segments of seq(17)
(
ð5Þ
Now, the similar operation was made with the sliding windows
on the 1,299 sequences in the testing dataset ST, and we obtained
250,208 positive samples and 573,791 negative samples, respec-
tively; i.e.,
Protein Domain Prediction
PLoS ONE | www.plosone.org2 June 2012 | Volume 7 | Issue 6 | e39308

Page 3

Sz
S{
Tcontains 250,208 positive segments
Tcontains 573,791 negative segments
(
ð6Þ
where Sz
ST, while S{
Trepresents the positive learning dataset derived from
Tthe corresponding negative dataset.
2. Feature Construction and Computational Method
2.1 The features of PSSM conservation scores.
a natural science with historic dimension. All biological species
have developed starting out from a very limited number of
ancestral species. The evolution in protein sequences involves
changes of single residues, insertions and deletions of several
residues [13], gene doubling, and gene fusion. In the course of
time such changes accumulate, so that many similarities between
initial and resultant amino acid sequences are eliminated, but the
corresponding proteins may still share many common attributes,
such as containing to a same domain and possessing basically the
same function. In view of this, evolutionary conservation will play
important roles in biological analysis: a more conserved residue
within a protein sequence may indicate that it is more important
for the protein function and thus under stronger selective pressure.
To incorporate this kind of evolutionary effects, we used PSSM
Biology is
(position-specific scoring matrix) [62] generated by Position
Specific Iterative BLAST (PSI BLAST) [63] to measure the
conservation status for a specific residue. A 20-dimensional vector
was used to denote the probabilities of conservation against
mutations to 20 different amino acids for a specific residue. For a
given sequence with L, its PSSM would correspond to a L|20
matrix, as formulated by equation 12 of [54]. Similar PSSM
approaches have been successfully used to enhance the prediction
qualityforvariousprotein
[21,26,27,50,54,55,56,64,65,66,67,68,69].
2.2 The features of amino acid factors.
20 amino acids has specific but different properties, the
composition of these properties of different residues within a
protein may have impacts on its structure and function. AAIndex
[70] is a database containing various physicochemical and
biochemical properties of amino acids. Atchley et al. [71]
performed multivariate statistical analyses on AAIndex and
transformed AAIndex to five multidimensional and highly
interpretable numeric patterns of attribute covariation that could
reflect (i) polarity, (ii) secondary structure, (iii) molecular volume,
(iv) codon diversity, and (v) electrostatic charge. Such five
numerical pattern scores, denoted as AAFactor (amino acid
factors), were used in this study to represent the respective
properties of each amino acid in a given protein.
attributes(see, e.g.,
Since each of the
Figure 1. A plot to show the change of the MCC values versus the feature numbers with different window sizes. The IFS curves were
drawn based on the data in Online Supporting Information S3. The MCC value reached the peak when the number of feature=360 and the window
size=13. The 360 features thus obtained were used to form the optimal feature set for the protein domain predictor. Purple line is for the case of
size-17 window, green for size-15 window, and brown for size-17 window. See the text for further explanation.
doi:10.1371/journal.pone.0039308.g001
Protein Domain Prediction
PLoS ONE | www.plosone.org3June 2012 | Volume 7 | Issue 6 | e39308

Page 4

2.3 The features of disorder score.
lacking fixed three-dimensional structures under physiological
conditions play important roles in biological functions [72,73,74].
The disordered regions of proteins allow for more modification
sites and interaction partners and always contain PTM (post
translational modification) sites, sorting signals, and protein
Protein segments
ligands. Therefore they are quite important for protein structure
and function [72,75,76]. In this study, the program VSL2 [77],
which can accurately predict both long and short disordered
regions in proteins, was used to calculate the disorder score that
denotes the disorder status of each amino acid in a given protein
sequence.
Table 1. The predicted results obtained with different window size.
Window
sizeDatasetSensitivityc<sen
0.577
Specificityc<spe
0.768
Accuracyc<acc
0.704
MCCc
13
SLa
0.342
STb
0.5780.7940.7280.367
15
SLa
0.570 0.7660.701 0.334
STb
0.571 0.7930.7260.360
17
SLa
0.5690.7670.701 0.333
STb
0.574 0.7930.726 0.362
a5-fold crossover test based on the learning dataset SL(cf. Eq.1).
bUsing the rule trained by SLto predict the query proteins in the independent dataset ST(cf. Eq.1).
cSee Eq.14 for more explanation.
doi:10.1371/journal.pone.0039308.t001
Figure 2. A 2-dimensional histogram to characterize the final optimal features set. The impact on the domain prediction from (A) the five
different feature types, and (B) each of the 13 subsites. See the text for further explanation.
doi:10.1371/journal.pone.0039308.g002
Protein Domain Prediction
PLoS ONE | www.plosone.org4 June 2012 | Volume 7 | Issue 6 | e39308

Page 5

2.4 The features of secondary structure and solvent
accessibility.
As is well known, the function of a protein is
closely correlated with its structure, and the post-translational
modification of specific residues may be affected by the solvent
accessibility of the relevant residues. Therefore, it would be useful
during the process of encoding the constituent amino acids by also
taking into account the features such as the secondary structure
propensity and solvent accessibility. These kinds of features could
be predicted by the software SSpro4 [78]. The second structural
propensity predicted by SSpro4 for each amino acid was ‘‘helix’’,
‘‘strand’’, or ‘‘other’’, encoded with ‘‘100’’, ‘‘010’’ and ‘‘001’’,
respectively; the solvent accessibility as ‘buried’ or ‘exposed’,
encoded with ‘‘10’’ and ‘‘01’’, respectively.
2.5 Feature space and feature vector.
in a given protein segment was formulated in terms of 31 features,
of which 20 from the PSSM scores, 1 from the disorder score, 5
from the AAFactor, 3 from the secondary structural propensities,
and 2 from the solvent accessibility states. Thus, each of the
segment samples generated by the size-13 sliding window would
contain 31|13~403 features; that by the size-15 sliding window,
31|15~465;andthat by
31|17~527. According to the general form of pseudo amino
acid composition (cf. equation 6 of [17], each of these segments
can be formulated by the following feature vector:
Each of the residues
thesize-17slidingwindow,
seq(j)~ y1
y2
???
yu
???
yV
½?T
ð7Þ
where yu(u~1,2,:::) represents the u-th feature score, T the
transpose operator, and
V
403,when j~13
465,when j~15
527, when j~17
8
>
>
:
<
ð8Þ
For those segments that contain the nominal residue ‘‘X’’, the
corresponding subsite was substituted with zero.
2.6 The mRMR method.
(minimal-redundancy-maximal-relevance) criterion [79] was used
to rank the importance of the features. The mRMR method could
rank the features according to their relevance to the target
concerned and the redundancy among the features themselves.
The ranked feature with a smaller index indicates that it has a
better trade-off between the maximum relevance and minimum
redundancy. To quantify both the relevance and redundancy, the
following mutual information (MI) is defined to estimate how one
vector is related to another:
In this study, the mRMR
Figure 3. A 2-dimensional histogram to characterize the PSSM features in the final optimal features set. (A) The impact on the domain
prediction from the mutation to each of the 20 amino acid types. (B) The evolutional conservation status for each of the 13 subsites. See the text for
further explanation.
doi:10.1371/journal.pone.0039308.g003
Protein Domain Prediction
PLoS ONE | www.plosone.org5June 2012 | Volume 7 | Issue 6 | e39308

Page 6

I(x,y)~
ð ð
p(x,y)log
p(x,y)
p(x)p(y)dxdy
ð9Þ
where x, y are two vectors, p(x,y) is the joint probabilistic density,
p(x) and p(y) are the marginal probabilistic densities. Suppose G
denotes the entire space containing all the feature components, Gs
denotes the already-selected feature set containing m features, and
Gtdenotes the to-be-selected feature set containing n features. The
relevance D between the feature f in G and the target c can be
calculated by
D~I(f,c)
ð10Þ
The redundancy R between the feature f in Gtand all the features
in Gscan be calculated by
R~1
m
X
fi[Gs
I(f,fi)
ð11Þ
To get the feature fj in Gt with the maximum relevance and
minimum redundancy, let us combine Eq.10 with Eq.11, as
formulated by
max
fj[Gt
I(fj,c){1
m
X
fi[Gs
I(fj,fi)
2
4
3
5
(j~1,2,???,n)
ð12Þ
The mRMR feature evaluation would continue N rounds when
given a feature set with N(~mzn) features. After these
evaluations, a feature set S can be obtained by the mRMR
method as formulated below
S~ f1
0,f2
0,???,fh
0,???,f
0
N
no
ð13Þ
where each feature in S has a subscript index indicating at which
round the feature is selected. The better the feature is, the earlier it
has been selected.
The mRMR program can be downloaded from the web-site at
http://penglab.janelia.org/proj/mRMR/.
2.7 The RF (random forest) method.
a popular machine-learning algorithm that has been recently
successfully used in dealing with various biological prediction
problems (see, e.g., [38,52,80,81,82,83]). Developed by Loe
Breiman [84], RF is an ensemble predictor consisting of multiple
decision trees. In Weka 3.6.4 [85], the classifier named with
‘‘RandomForest’’ has implemented the predictor. In the current
The RF approach is
Figure 4. A 2-dimensional histogram to characterize the amino acid factor types in the final optimal features set. The impact on the
domain prediction from (A) the five different amino acid types, and (B) each of the 13 subsites. See the text for further explanation.
doi:10.1371/journal.pone.0039308.g004
Protein Domain Prediction
PLoS ONE | www.plosone.org 6June 2012 | Volume 7 | Issue 6 | e39308

Page 7

study, RandomForest was adopted as the prediction engine and
operated with the default parameters. During the process of
classifying a queried sample with its feature vector, RandomForest
first grew 10 decision trees according to the following procedures.
(i) Suppose the number of training cases is N, take N samples at
random – but with replacement, from the original data. These
samples are to form the training set for growing the tree. Here the
so-called ‘‘with replacement’’ is a mathematical term meaning that
a sample selected at random from the original dataset is returned
to the original dataset before a second one is selected at random.
In other words, whenever a sample is selected, the original dataset
contains all the same samples. Thus, an exactly same sample may
be selected more than once, and there is no change at all in the
size of the original dataset at any stage. (ii) If each case consists of
M input features, choose a number m=[log2M+1] which is much
less than M. At each node, m features are selected randomly out of
the M features and the most optimized split on these m features is
employed to split the node. The value of m does not change during
the growth of the tree. (iii) Each tree is fully grown and not pruned.
Then the input vector is predicted by each of 10 decision tree and
10 predicted classes provided by them are obtained. Finally, the
class with the most votes will be selected as the output class of
RandomForest.
The Weka program package can be downloaded from the web-
site at http://www.cs.waikato.ac.nz/ml/weka/index_downloading.
html
2.8 The cross-validation method.
the following three cross-validation methods are often used to
examine a predictor for its effectiveness in practical application:
independent dataset test, subsampling test, and jackknife test [86].
However, as elucidated in [58] and demonstrated by Eqs.28–32 of
[17], among the three cross-validation methods, the jackknife test
is deemed the least arbitrary (most objective) that can always yield
a unique result for a given benchmark dataset, and hence has been
increasingly used and widely recognized by investigators to
examinetheaccuracy ofvarious
[20,26,27,87,88,89,90,91]). However, to reduce the computational
time, we adopted the 5-fold cross-validation in this study as done
by many investigators with SVM as the prediction engine (see, e.g.,
[92,93,94]). During the process of 5-fold cross-validation, the
benchmark dataset was first equally divided into 5 subsets.
Subsequently, each of the subsets was in turn used as the testing
dataset and the remaining four subsets as the training or learning
dataset. To evaluate the performance of the predictor, the
prediction accuracy, specificity, sensitivity and MCC (Matthews’s
correlation coefficient) were calculated below:
In statistical prediction,
predictors(see,e.g.,
Figure 5. A 2-dimensional histogram to characterize the solvent accessibility types in the final optimal features set. The impact on the
protein domain prediction from (A) the two different types of the solvent accessibility, and (B) each of the 13 subsites. See the text for further
explanation.
doi:10.1371/journal.pone.0039308.g005
Protein Domain Prediction
PLoS ONE | www.plosone.org7 June 2012 | Volume 7 | Issue 6 | e39308

Page 8

<sen~
TP
TPzFN
TN
TNzFP
<spe~
<aac~
TPzTN
TPzFPzTNzFN
MCC~
(TP)(TN)-(FP)(FN)
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
where <senreflects the sensitivity, i.e., the rate of positive samples
that are correctly predicted as positive; <spereflects the specificity,
i.e., the rate of negative samples that are correctly predicted as
negative; <acc reflects the accuracy, i.e., the rate of correctly
predicted events; MCC is the Matthew’s correlation coefficient;
TP represents the true positive; TN, the true negative; FP, the false
positive; and FN, the false negative.
2.9TheIFS(incremental
approach.
Based on the ranked features according to their
importance evaluated by the mRMR approach, we used the IFS
[95,96,97] approach to determine the optimal number of features.
During the IFS procedure, features in the ranked feature set were
added with a stepwise of l from higher to lower rank. A new
½TPzFP?½TPzFN?½TNzFP?½TNzFN?
p
8
>
>
>
>
>
>
>
>
>
>
>
>
>
>
:
>
<
>
>
>
>
>
>
>
ð14Þ
featureselection)
feature set was formed when l features had been added. Thus
N=l
½
th feature set is:
? feature sets would be composed for N ranked features. The i-
Si~fSl,S2l,???,Silg
(1ƒiƒ½N=l?)
ð15Þ
where N denotes the total number of features in the original
dataset and l (step) is a positive integer. In this study l~5. For each
of the [N/l] feature sets, an RF classifier was constructed and
examined using the 5-fold cross-validation on the benchmark
dataset. By doing so we obtained an IFS table with one column for
the index i and the other four columns for the prediction accuracy,
sensitivity, specificity and MCC, respectively. Thus, we could
obtain the optimal feature set (Soptimal), with which the predictor
would yield the best prediction performance.
2.10 The final optimal feature set.
fluctuating with the increase of feature numbers. Therefore, it was
necessary to carefully examine its variation against the increasing
feature number. In this study the feature-increasing gap was set at
5 to winnow out the optimal features. In other words, we
compared two neighbor MCC values at a time with a stepwise of
five features, if the latter MCC value is greater than the former
one, then the corresponding five features were reserved to join the
The MCC curve was
Figure 6. A 2-dimensional histogram to characterize the secondary structure types in the final optimal features set. The impact on the
domain prediction from (A) the three different secondary structure types, and (B) each of the 13 subsites. See the text for further explanation. See the
text for further explanation.
doi:10.1371/journal.pone.0039308.g006
Protein Domain Prediction
PLoS ONE | www.plosone.org8June 2012 | Volume 7 | Issue 6 | e39308

Page 9

final optimal feature set; otherwise, discarded. The final optimal
feature set thus established consisted of 195 features and would be
used for further analysis.
We installed Weka into our Linux machine. Its ‘‘Run
Environment and Configuration’’ was: Hardware 2 Intel(R)
Xeon(R); CPU E5335@2.00 GHz; 16 G RAM; OS CentOS
release 4.9 (Final) x86_64.
Figure 7. Illustration to show the predicted results obtained before and after applying the sequence-scanning refinement
operation. A residue assigned to the domain region was coded with ‘‘1’’; otherwise, ‘‘2’’. The 3D structure of A1A5Q6 was retrieved from ModBase.
See the text for further explanation.
doi:10.1371/journal.pone.0039308.g007
Protein Domain Prediction
PLoS ONE | www.plosone.org9 June 2012 | Volume 7 | Issue 6 | e39308

Page 10

Results and Discussion
1. The mRMR Result
Listed in the Online Supporting Information S2 are two kinds of
outcomes obtained by running the mRMR software: one is called
the ‘‘MaxRel feature list’’ that ranked all the features according to
their relevance to the class of samples; the other one is the
‘‘mRMR feature list’’ that ranked the features according to the
criteria of maximum relevance and minimum redundancy. In the
mRMR feature list, the smaller the index of a feature was, the
more important the feature would be for the protein domain
prediction. Accordingly, the mRMR feature list could be used to
establish the optimal feature set in the IFS procedure.
2. IFS and Final Optimal Feature Set
In Section 2.9 of Materials and Methods, by setting 403 for N
and 5 for the feature-increasing gap, 80 individual predictors
corresponding to 80 feature subsets were established for predicting
the protein domain sites in the sequence samples generated by the
size-13 sliding window. Listed in the Online Supporting Informa-
tion S3 are the rates of prediction accuracy, specificity, sensitivity
and MCC (cf. 14) obtained by each of the 80 predictors. Shown in
Fig. 1 is the IFS curve plotted based on the data in Online
Supporting Information S3. The same calculations were also
carried out for the size-15 and size-17 windows, and the
corresponding results were also plotted in Fig. 1, from which
we can see that the predictor based on the size-13 window
outperformed the other two, and that the maximal MCC was
0.342 when 360 features were included. These 360 features were
deemed to form the optimal feature set of our classifier. With such
a classifier, the prediction sensitivity, specificity and accuracy were
0.577, 0.768 and 0.704 respectively (Table 1). The optimal 360
features were given in the Online Supporting Information S4.
After taking the IFS procedure (cf. In Sections 2.9 and 2.10 of
Materials and Methods), we obtained the 195 final optimal
features as given in the Online Supporting Information S5.
Hereafter, all the analyses will be based on such 195 final
optimal features.
The CPU time of the above calculation for size-13, 15 and 17
windows were about 4 hours, 5 hours and 6 hours, respectively.
3. Feature Analysis
The distribution of the number of each type of features in the
final optimal feature set was investigated and shown in Fig. 2A. Of
the 195 optimal features, 147 were from PSSM conservation
scores, 21 from the amino acid factors, 4 from the disorder scores,
7 from the solvent accessibilities, and 16 from the secondary
structural propensities. All these five kinds of features made
contributions to the prediction of protein domain sites. It was
revealed by the site-specific distribution of the optimal feature set
(see Fig. 2B) that sites 1–2, site 10 and site 13 played most
important roles in determining the domain sites. In addition, the
features of site 4 and site 5 also had considerable impacts on the
prediction of protein domain sites.
4. PSSM Conservation Score Feature Analysis
As mentioned above, among the 195 optimal features, 147
belonged to the PSSM conservation features and hence had the
highest proportion. It can be clearly seen from Fig. 3A that each
of the 20 different amino acid types would have different PSSM
conservation impact in determining the protein domain site. In
this regard, the amino acid N (asparagine) or D (aspartic acid)
would have the highest impact, successively followed by G
(glycine), R (arginine), and so forth. Interestingly, it has been
reported that D, G and R were over-represented in protein
interaction domains [98]. Besides, G was believed to be
instrumental in defining the core domain and inter-domain
regions of a protein [39]. Meanwhile, as shown in Fig. 3B, for
the samples generated by the size-13 window (cf. Eq.2), the
conservation status at the subsite 10 played the most important
role in predicting the protein domain site, followed by the subsites
1, 2, 4, and 7. Furthermore, of the top ten features in the final
optimal feature list, five were from the PSSM conservation
features. The first one was the conservation status against residue
M (methionine) at subsite 1 (index 3, ‘‘AA1_pssm_13’’). The
other four were the conservation status against residue A at
subsite 12 (index 4, ‘‘AA12_pssm_1’’), the conservation status
against residue G at subsite 6 and site 4 (index 6 and index 7,
‘‘AA6_pssm_8’’ and AA4_pssm_8), and the conservation status
Table 2. A comparison between the predicted results with and without the scanning refinement.
Window sizeScanning refinementa
Sensitivityb<sen
0.578
Specificityb<spe
0.794
Accuracyb<acc
0.728
MCCb
13 No 0.367
Yes 0.6420.808 0.7580.441
15No0.5710.7930.7260.360
Yes 0.6340.8060.7540.431
17 No0.574 0.7930.7260.362
Yes0.6450.804 0.7560.438
aSee section 9 of Results and Discussion for more explanation about the scanning procedure.
bSee Eq.14 for more explanation.
doi:10.1371/journal.pone.0039308.t002
Table 3. Comparison of the current method with the existing
methods on the same testing dataset ST(cf. Eq.1).
Method
Sensitivity
<sen
0.643
Specificity
<spe
0.808
Accuracy
<acc
0.757
MCC
Our method0.441
DoMpro [40] 0.9240.1820.406 0.138
Globplot [49]0.8680.325 0.4850.199
Domcut [48]a
0.9790.110 0.367 0.149
Domcut [48]b
0.856 0.3250.4820.186
aUsing the default cutoff threshold of 20.09.
bUsing the optimal cutoff threshold,
doi:10.1371/journal.pone.0039308.t003
Protein Domain Prediction
PLoS ONE | www.plosone.org 10June 2012 | Volume 7 | Issue 6 | e39308

Page 11

against
‘‘AA2_pssm_17’’)
residueT(threonine) atsubsite2 (index8,
5. Amino Acid Factor Analysis
Illustrated in Fig. 4 are the impacts of different amino acid
factors and their subsite locations to the protein domain
prediction. It can be seen from Fig. 4A that the codon diversity
was the most important feature to the protein domain site
prediction, as supported by [98,99]. Besides, it was reported that
‘‘codon harmonization’’ would put some non-preferred codons
into the positions corresponding to the predicted protein domain
boundaries [100]. Furthermore, the electrostatic charge has
proved to be essential for the localization and activation of many
proteins containing polycationic domains in their amino acid
sequence [101]. Meanwhile, it has also been revealed that
binding of oppositely charged proteins via electrostatic interac-
tions can induce domain formation [102]. As shown in Fig. 4B,
the amino acid residues at the subsite 2 and site 13 would have
the highest impact to the protein domain sites prediction.
Interestingly, the electrostatic feature at the subsite 13 had an
index of 2 in our final optimal feature set, indicating that it was
one of the most important features for the protein domain site
prediction.
6. Disorder Analysis
Within the final optimal feature set, four of all the 13 disorder
features were selected, indicating that the disorder feature might
play a pivotal role in protein domain site prediction. Such four
disorder features were from subsites 1, 5, 10 and 13. Particularly,
the disorder feature of subsite 5 had the index of 1 in the final
optimal feature set, suggesting that it was the most important
feature in the protein domain site prediction. Also, the disorder
feature of subsite 13 has an index of 9 in the final optimal feature
site. These findings are fully consistent with the observations that
the regions of substantial structural flexibility in a protein often
correspond to domain boundaries where the structure is usually
exposed and less constrained [39].
7. Solvent Accessibility Features Analysis
Shown in Fig. 5 are the solvent accessibility features in the
optimal feature set. It can be seen from Fig. 5A that the number
of buried solvent accessibility features was much more than that of
the exposed, indicating that the protein domains were skewed
toward the buried areas. Such findings are consistent with the
report the buried protein regions can be accessible to water when
they are in a free subunit or in one domain state and can form a
complex or an aggregate with other subunits or domains [103].
Moreover, it can be seen from Fig. 5B that the solvent
accessibility features at the subsites 2, 3, 8, 9, and 11–13 have
relatively more impacts on the domain site prediction.
8. Secondary Structure Features Analysis
The feature and site-specific distribution of the secondary
structure in the optimal feature set was given in Fig. 6, from which
we can see that the features of ‘‘strand’’ and ‘‘other’’ did affect the
domain site prediction (panel A), while the secondary structure
features at subsites 1, 5, 6, 8 and 13 had relatively more impact on
the domain site determination (panel B).
9. Scan the Entire Protein Sequence to Refine the Domain
Region Prediction
As mentioned above, each of the amino acid residues in a
protein sequence was identified whether it belonged to a domain
region or non-domain region (cf. Eq.2). If a residue was identified
as belonging to a domain region, it was coded with ‘‘1’’; otherwise,
‘‘2’’, as illustrated in Fig. 7. However, it is inevitable that some
domain residues might be mispredicted as non-domain residues
resulting in some short strand of ‘‘2’’ inserted in a long strand of
‘‘1’’ and vice versa. To filter out this kind of false positives and false
negatives, a special scanning algorithm was developed to refine the
entire predicted results according to the following criteria. (i) Any
negative code ‘‘2’’ should be modified to a positive code ‘‘1’’ if it
followed a strand of more than four continuous ‘‘1’’ codes but was
followed by less than four continuous ‘‘2’’ codes. (ii) Any positive
code ‘‘1’’ should be modified to a negative code ‘‘2’’ if it followed a
strand of more than four continuous ‘‘2’’ codes but was followed
by less than three continuous ‘‘1’’ codes. After such a scanning
procedure, it can be seen from Fig. 7 that many sporadic ‘‘2’’
codes in the long ‘‘1’’ regions have disappeared, and that many
sporadic ‘‘1’’ codes in the long ‘‘2’’ regions have disappeared too.
Meanwhile, the prediction quality was further improved as
indicated in Table 2. Finally, the regions with the long continuous
‘‘1’’ codes thus obtained were assigned corresponding to the
domain regions as indicated in Online Supporting Information S6.
10. In Comparison with the Existing Methods
To evaluate our method, let us compare its performance with
three existing methods in this area, including DoMpro [40],
Globplot [49] and Domcut [48] based on the same testing dataset.
Those methods such as FIEFDom [47] were not included because
they were aimed at predicting domain boundaries rather than
domains themselves. In other word, this kind of methods was
based on such an assumption that nearly the whole protein was
domain region except two or three domain boundaries. As a
consequence, their sensitivity <senwould be very close to 1, but the
specificity <spewould be very low with quite poor overall success
rates. The prediction result by the DoMpro [40] on a query
protein sequence was formulated by a series of ‘‘N’’ and ‘‘T’’ codes
to indicate that the corresponding residue being outside and inside
the domain region, respectively. The predicted outcomes by the
Globplot method [49] were the domain regions directly. As for the
method Domcut [48], a score was assigned to each of the
constituent residues in a query protein. The residues with a score
below the cutoff threshold (default 20.09) were regarded as the
inter-domain linker regions. For facilitating comparison, the results
by all these methods on the same independent dataset ST(cf.
Eq.1) are also listed in Table 3, from which we can see that our
method was about 58–70% higher than the other methods in
specificity, 28–40% higher in accuracy, and 24–31% higher in
MCC, but about 20% lower in sensitivity. These results indicate
that the current method will play an important complementary
role to the existing methods in identifying the domains of proteins.
11. Useful Insights for Guiding Experiments or Being
Validated by Experiments
The selected features at different sites may provide clues or
insights for researchers to find or validate new protein domains, as
can be viewed from the following four aspects. (i) PSSM feature. It
was found through analyzing the PSSM conservation score that
the mutations to amino acid residues N and D had the most
impact on identifying the protein domain sites. Besides, the
mutation to residues G and R also had more impacts than the
other amino acids in this regard, fully consistent with the report
[98] that D, G and R were over-represented in protein interaction
domains, and the report [39] that amino acid G was instrumental
in defining the core domain and interdomain regions of a protein.
Protein Domain Prediction
PLoS ONE | www.plosone.org11 June 2012 | Volume 7 | Issue 6 | e39308

Page 12

(ii) Codon diversity feature. It was revealed in this study that the
codon diversity played pivotal role in identifying the protein
domain sites, as evidenced by a series of experiments [98,99,100].
(iii) Electrostatic charge feature. It is interesting to note that
electrostatic charge has proved to be essential for the localization
and activation of many proteins containing polycationic domains
in their amino acid sequence [101], and that binding of oppositely
charged proteins via electrostatic interactions can induce domain
formation [102]. All these observations are quite consistent with
the findings in this study that the electrostatic feature of site 13 has
an index of 2 in our final optimal feature set meaning that it is one
of the most important features for the protein domain sites
prediction. (iv) Disorder feature. It was found that in the final
optimal feature set derived from this study, four of all the 13
disorder features were selected, and that disorder feature of site 5
had the index of 1, implying it was the most important feature to
the protein domain site prediction. Interestingly, it has been
reported that disorder regions often correspond to the domain
boundaries [39]. Accordingly, the remainders in the optimal
feature set are certainly worth being further investigated by future
experiments.
It is anticipated that the strategy and approaches developed in
this study may also be extended to investigate protein signal
peptides (see, e.g., [60,61,104,105]), B-cell epitopes [106,107],
HIV protease cleavage sites [108,109,110,111], enzyme specificity
[112,113], among many other important topics in protein science
and biomedicine.
Supporting Information
Supporting Information S1
proteins in the benchmark dataset: (A) training dataset and (B)
testing dataset.
(XLSX)
The UniProt ID codes for the
Supporting Information S2
mRMR feature selection and analysis operation. It contains two
sheets. The first one shows the 403 maximum relevance features of
the size-13 window ranked according to their relevance with the
Detailed results obtained by the
classification of the samples. The second one shows the 403
mRMR features of the size-13 window ranked according to the
redundancy and relevance criteria.
(XLSX)
Supporting Information S3
specificity (Sp), accuracy (Ac), and Matthews’s correlation
coefficient (MCC) versus the feature numbers. It contains three
sheets. The first one shows the IFS results for the size-13 window,
the second for the size-15 window, and the third for the size-17
window. The IFS curves were plotted based on the data in this file.
See the text for more explanation.
(XLSX)
The results of the sensitivity (Sn),
Supporting Information S4
IFS procedure for the size-13 window. See the text for more
explanation.
(XLSX)
The 360 features selected by the
Supporting Information S5
selected by the incremental analysis procedure for the size-13
window. See the text for more explanation.
(XLSX)
The 195 final optimal features
Supporting Information S6
proteins predicted by the method proposed in this paper.
(XLSX)
Domain region of the 1,299
Acknowledgments
The authors wish to thank the editor for taking time to edit this paper. The
authors would also like to thank the two anonymous reviewers for their
constructive comments, which were very helpful for strengthening the
presentation of this study.
Author Contributions
Conceived and designed the experiments: BQL LLH YDC. Performed the
experiments: BQL LLH. Analyzed the data: BQL KCC. Contributed
reagents/materials/analysis tools: BQL LC. Wrote the paper: BQL KYF
KCC.
References
1. Chivian D, Kim DE, Malmstrom L, Bradley P, Robertson T, et al. (2003)
Automated prediction of CASP-5 structures using the Robetta server. Proteins
53 Suppl 6: 524–533.
2. Ingolfsson H, Yona G (2008) Protein domain prediction. Methods Mol Biol
426: 117–143.
3. Holland TA, Veretnik S, Shindyalov IN, Bourne PE (2006) Partitioning protein
structures into domains: why is it so difficult? J Mol Biol 361: 562–590.
4. Campbell ID, Downing AK (1994) Building protein structure and function
from modular units. Trends Biotechnol 12: 168–172.
5. Guerois R, Serrano L (2001) Protein design based on folding models. Curr
Opin Struct Biol 11: 101–106.
6. Nielsen PK, Yamada Y (2001) Identification of cell-binding sites on the
Laminin alpha 5 N-terminal domain by site-directed mutagenesis. J Biol Chem
276: 10906–10912.
7. Chou KC (2004) Review: Structural bioinformatics and its impact to
biomedical science. Current Medicinal Chemistry 11: 2105–2134.
8. Schnell JR, Chou JJ (2008) Structure and mechanism of the M2 proton channel
of influenza A virus. Nature 451: 591–595.
9. Wang J, Pielak RM, McClintock MA, Chou JJ (2009) Solution structure and
functional analysis of the influenza B proton channel. Nature Structural and
Molecular Biology 16: 1267–1271.
10. Chou JJ, Li S, Klee CB, Bax A (2001) Solution structure of Ca2+-calmodulin
reveals flexible hand-like properties of its domains. Nature Structural Biology 8:
990–997.
11. Berardi MJ, Shih WM, Harrison SC, Chou JJ (2011) Mitochondrial uncoupling
protein 2 structure determined by NMR molecular fragment searching. Nature
476: 109–113.
12. Chou KC (2004) Insights from modelling the 3D structure of the extracellular
domain of alpha7 nicotinic acetylcholine receptor. Biochemical and Biophys-
ical Research Communication 319: 433–438.
13. Chou KC (1995) The convergence-divergence duality in lectin domains of the
selectin family and its implications. FEBS Letters 363: 123–126.
14. Chou KC (2004) Modelling extracellular domains of GABA-A receptors:
subtypes 1, 2, 3, and 5. Biochemical and Biophysical Research Communica-
tions 316: 636–642.
15. Chou KC (2005) Modeling the tertiary structure of human cathepsin-E.
Biochem Biophys Res Commun 331: 56–60.
16. Chou KC (2001) Prediction of protein cellular attributes using pseudo amino
acid composition. PROTEINS: Structure, Function, and Genetics (Erratum:
ibid, 2001, Vol44, 60) 43: 246–255.
17. Chou KC (2011) Some remarks on protein attribute prediction and pseudo
amino acid composition (50th Anniversary Year Review). Journal of
Theoretical Biology 273: 236–247
18. Cai YD, Zhou GP, Chou KC (2003) Support vector machines for predicting
membrane protein types by using functional domain composition. Biophysical
Journal 84: 3257–3263.
19. Xiao X, Wang P, Chou KC (2009) GPCR-CA: A cellular automaton image
approach for predicting G-protein-coupled receptor functional classes. Journal
of Computational Chemistry 30: 1414–1423.
20. Xiao X, Wang P, Chou KC (2011) GPCR-2L: Predicting G protein-coupled
receptors and their types by hybridizing two different modes of pseudo amino
acid compositions. Molecular Biosystems 7: 911–919.
21. Chou KC, Shen HB (2008) ProtIdent: A web server for identifying proteases
and their types by fusing functional domain and sequential evolution
information. Biochem Biophys Res Comm 376: 321–325.
22. Xiao X, Wang P, Chou KC (2011) Quat-2L: a web-server for predicting
protein quaternary structural attributes. Molecular Diversity 15: 149–155.
23. Xiao X, Wang P, Chou KC (2009) Predicting protein quaternary structural
attribute by hybridizing functional domain composition and pseudo amino acid
composition. Journal of Applied Crystallography 42: 169–173.
Protein Domain Prediction
PLoS ONE | www.plosone.org12 June 2012 | Volume 7 | Issue 6 | e39308

Page 13

24. Chou KC, Cai YD (2004) Predicting protein structural class by functional
domain composition. Biochemical and Biophysical Research Communications
(Corrigendum: ibid, 2005, Vol329, 1362) 321: 1007–1009.
25. Wang K, Hu LL, Shi XH, Dong YS, Li HP, et al. (2012) PSCL: Predicting
Protein Subcellular Localization Based on Optimal Functional Domains.
Protein & Peptide Letters 19: 15–22.
26. Chou KC, Shen HB (2010) A new method for predicting the subcellular
localization of eukaryotic proteins with both single and multiple sites: Euk-
mPLoc 2.0. PLoS ONE 5: e9931.
27. Chou KC, Shen HB (2010) Plant-mPLoc: A Top-Down Strategy to Augment
the Power for Predicting Plant Protein Subcellular Localization. PLoS ONE 5:
e11335.
28. Zhou GP (2011) The Structural Determinations of the Leucine Zipper Coiled-
Coil Domains of the cGMP-Dependent Protein Kinase I alpha and its
Interaction with the Myosin Binding Subunit of the Myosin Light Chains
Phosphase. Proteins & Peptide Letters 18: 966–978.
29. Zhou GP (2011) The disposition of the LZCC protein residues in wenxiang
diagram provides new insights into the protein-protein interaction mechanism.
Journal of Theoretical Biology 284: 142–148.
30. Chen L, Feng KY, Cai YD, Chou KC, Li HP (2010) Predicting the network of
substrate-enzyme-product triads by combining compound similarity and
functional domain composition. BMC Bioinformatics 11: 293.
31. Gewehr JE, Zimmer R (2006) SSEP-Domain: protein domain prediction by
alignment of secondary structure elements and profiles. Bioinformatics 22:
181–187.
32. von Ohsen N, Sommer I, Zimmer R, Lengauer T (2004) Arby: automatic
protein structure prediction using profile-profile alignment and confidence
measures. Bioinformatics 20: 2228–2235.
33. Zhang Y, Liu B, Dong Q, Jin VX (2011) An improved profile-level domain
linker propensity index for protein domain boundary prediction. Protein &
Peptide Letters 18: 7–16.
34. George RA, Heringa J (2002) SnapDRAGON: a method to delineate protein
structural domains from sequence data. J Mol Biol 316: 839–851.
35. Liu J, Rost B (2004) Sequence-based prediction of protein domains. Nucleic
Acids Res 32: 3522–3530.
36. Kim DE, Chivian D, Malmstrom L, Baker D (2005) Automated prediction of
domain boundaries in CASP6 targets using Ginzu and RosettaDOM. Proteins
61 Suppl 7: 193–200.
37. Cheng J (2007) DOMAC: an accurate, hybrid protein domain prediction
server. Nucleic Acids Res 35: W354–356.
38. Shameer K, Pugalenthi G, Kandaswamy KK, Sowdhamini R (2011) 3dswap-
pred: Prediction of 3D Domain Swapping from Protein Sequence Using
Random Forest Approach. Protein & Peptide Letters 18: 1010–1020.
39. Nagarajan N, Yona G (2004) Automatic prediction of protein domains from
sequence information using a hybrid learning system. Bioinformatics 20: 1335–
1360.
40. Cheng J, Sweredoski M, Baldi P (2006) DOMpro: Protein Domain Prediction
Using Profiles, Secondary Structure, Relative Solvent Accessibility, and
Recursive Neural Networks. Data Mining and Knowledge Discovery 13: 1–10.
41. Eickholt J, Deng X, Cheng J (2011) DoBo: Protein domain boundary
prediction by integrating evolutionary signals and machine learning. BMC
Bioinformatics 12: 43.
42. Ebina T, Toh H, Kuroda Y (2011) DROP: an SVM domain linker predictor
trained with optimal features selected by random forest. Bioinformatics 27:
487–494.
43. Orengo CA, Michie AD, Jones S, Jones DT, Swindells MB, et al. (1997)
CATH–a hierarchic classification of protein domain structures. Structure 5:
1093–1108.
44. Murzin AG, Brenner SE, Hubbard T, Chothia C (1995) SCOP: a structural
classification of proteins database for the investigation of sequences and
structures. J Mol Biol 247: 536–540.
45. Holm L, Sander C (1998) Dictionary of recurrent domains in protein
structures. Proteins 33: 88–96.
46. Walsh I, Martin AJ, Mooney C, Rubagotti E, Vullo A, et al. (2009) Ab initio
and homology based prediction of protein domains by recursive neural
networks. BMC Bioinformatics 10: 195.
47. Bondugula R, Lee MS, Wallqvist A (2009) FIEFDom: a transparent domain
boundary recognition system using a fuzzy mean operator. Nucleic Acids Res
37: 452–462.
48. Suyama M, Ohara O (2003) DomCut: prediction of inter-domain linker
regions in amino acid sequences. Bioinformatics 19: 673–674.
49. Linding R, Russell RB, Neduva V, Gibson TJ (2003) GlobPlot: Exploring
protein sequences for globularity and disorder. Nucleic Acids Res 31: 3701–
3708.
50. Wu ZC, Xiao X, Chou KC (2011) iLoc-Plant: a multi-label classifier for
predicting the subcellular localization of plant proteins with both single and
multiple sites. Molecular BioSystems 7: 3287–3297.
51. Xiao X, Wu ZC, Chou KC (2011) A multi-label classifier for predicting the
subcellular localization of gram-negative bacterial proteins with both single and
multiple sites. PLoS ONE 6: e20592.
52. Lin WZ, Fang JA, Xiao X, Chou KC (2011) iDNA-Prot: Identification of DNA
Binding Proteins Using Random Forest with Grey Model. PLoS ONE 6:
e24756.
53. Wang P, Xiao X, Chou KC (2011) NR-2L: A Two-Level Predictor for
Identifying Nuclear Receptor Subfamilies Based on Sequence-Derived
Features. PLoS ONE 6: e23505.
54. Chou KC, Wu ZC, Xiao X (2012) iLoc-Hum: Using accumulation-label scale
to predict subcellular locations of human proteins with both single and multiple
sites. Molecular Biosystems 8: 629–641.
55. Xiao X, Wu ZC, Chou KC (2011) iLoc-Virus: A multi-label learning classifier
for identifying the subcellular localization of virus proteins with both single and
multiple sites. Journal of Theoretical Biology 284: 42–51.
56. Wu ZC, Xiao X, Chou KC (2012) iLoc-Gpos: A Multi-Layer Classifier for
Predicting the Subcellular Localization of Singleplex and Multiplex Gram-
Positive Bacterial Proteins. Protein & Peptide Letters 19: 4–14.
57. Apweiler R, Martin MJ, O’Donovan C, Magrane M, Alam-Faruque Y, et al.
(2010) The Universal Protein Resource (UniProt) in 2010. Nucleic Acids
Research 38: D142–D148.
58. Chou KC, Shen HB (2007) Review: Recent progresses in protein subcellular
location prediction. Analytical Biochemistry 370: 1–16.
59. Li W, Godzik A (2006) Cd-hit: a fast program for clustering and comparing
large sets of protein or nucleotide sequences. Bioinformatics 22: 1658–1659.
60. Chou KC (2002) Review: Prediction of protein signal sequences. Current
Protein and Peptide Science 3: 615–622.
61. Chou KC, Shen HB (2007) Signal-CF: a subsite-coupled and window-fusing
approach for predicting signal peptides. Biochem Biophys Res Comm 357:
633–640.
62. Schaffer AA, Aravind L, Madden TL, Shavirin S, Spouge JL, et al. (2001)
Improving the accuracy of PSI-BLAST protein database searches with
composition-based statistics and other refinements. Nucleic Acids Res 29:
2994–3005.
63. Altschul SF, Madden TL, Schaffer AA, Zhang J, Zhang Z, et al. (1997) Gapped
BLAST and PSI-BLAST: a new generation of protein database search
programs. Nucleic Acids Res 25: 3389–3402.
64. Chou KC, Shen HB (2007) MemType-2L: A Web server for predicting
membrane proteins and their types by incorporating evolution information
through Pse-PSSM. Biochem Biophys Res Comm 360: 339–345.
65. Hayat M, Khan A (2012) MemHyb: Predicting membrane protein types by
hybridizing SAAC and PSSM. J ournal of Theoretical Biology 292: 93–102.
66. Li D, Jiang Z, Yu W, Du L (2010) Predicting Caspase Substrate Cleavage Sites
Based on a Hybrid SVM-PSSM Method. Protein and Peptide Letters 17:
1566–1571.
67. Mundra P, Kumar M, Kumar KK, Jayaraman VK, Kulkarni BD (2007) Using
pseudo amino acid composition to predict protein subnuclear localization:
Approached with PSSM. Pattern Recognition Letters 28: 1610–1615.
68. Shen HB, Chou KC (2007) Nuc-PLoc: A new web-server for predicting protein
subnuclear localization by fusing PseAA composition and PsePSSM. Protein
Engineering, Design & Selection 20: 561–567.
69. Chou KC, Wu ZC, Xiao X (2011) iLoc-Euk: A Multi-Label Classifier for
Predicting the Subcellular Localization of Singleplex and Multiplex Eukaryotic
Proteins. PLoS One 6: e18258.
70. Kawashima S, Kanehisa M (2000) AAindex: amino acid index database.
Nucleic Acids Res 28: 374.
71. Atchley WR, Zhao J, Fernandes AD, Druke T (2005) Solving the protein
sequence metric problem. Proc Natl Acad Sci U S A 102: 6395–6400.
72. Wright PE, Dyson HJ (1999) Intrinsically unstructured proteins: re-assessing
the protein structure-function paradigm. Journal of Molecular Biology 293:
321–331.
73. Dunker AK, Brown CJ, Lawson JD, Iakoucheva LM, Obradovic Z (2002)
Intrinsic disorder and protein function. Biochemistry 41: 6573–6582.
74. Yoon MK, Venkatachalam V, Huang A, Choi BS, Stultz CM, et al. (2009)
Residual structure within the disordered C-terminal segment of p21(Waf1/
Cip1/Sdi1) and its implications for molecular recognition. Protein Sci 18: 337–
347.
75. Liu J, Tan H, Rost B (2002) Loopy proteins appear conserved in evolution.
Journal of Molecular Biology 322: 53–64.
76. Tompa P (2002) Intrinsically unstructured proteins. Trends in Biochemical
Sciences 27: 527–533.
77. Peng K, Radivojac P, Vucetic S, Dunker AK, Obradovic Z (2006) Length-
dependent prediction of protein intrinsic disorder. BMC Bioinformatics 7: 208.
78. Cheng J, Randall AZ, Sweredoski MJ, Baldi P (2005) SCRATCH: a protein
structure and structural feature prediction server. Nucleic Acids Research 33:
W72–W76.
79. Peng H, Long F, Ding C (2005) Feature selection based on mutual information:
criteria of max-dependency, max-relevance, and min-redundancy. IEEE Trans
Pattern Anal Mach Intell 27: 1226–1238.
80. Kandaswamy KK, Chou KC, Martinetz T, Moller S, Suganthan PN, et al.
(2011) AFP-Pred: A random forest approach for predicting antifreeze proteins
from sequence-derived properties. Journal of Theoretical Biology 270: 56–62.
81. Pugalenthi G, Kandaswamy KK, Chou KC, Vivekanandan S, Kolatkar P
(2012) RSARF: Prediction of Residue Solvent Accessibility from Protein
Sequence Using Random Forest Method. Protein & Peptide Letters 19: 50–56.
82. Jia SC, Hu XZ (2011) Using Random Forest Algorithm to Predict beta-Hairpin
Motifs. Protein and Peptide Letters 18: 609–617.
83. Qiu Z, Wang X (2011) Improved Prediction of Protein Ligand-Binding Sites
Using Random Forests. Protein & Peptide Letters 18: 1212–1218.
84. Breiman L (2001) Random forests. Machine learning 45: 5–32.
Protein Domain Prediction
PLoS ONE | www.plosone.org 13 June 2012 | Volume 7 | Issue 6 | e39308

Page 14

85. Witten IH, Frank E (2005) Data Mining: Practical machine learning tools and
techniques: Morgan Kaufmann Pub.
86. Chou KC, Zhang CT (1995) Review: Prediction of protein structural classes.
Critical Reviews in Biochemistry and Molecular Biology 30: 275–349.
87. Esmaeili M, Mohabatkar H, Mohsenzadeh S (2010) Using the concept of
Chou’s pseudo amino acid composition for risk type prediction of human
papillomaviruses. Journal of Theoretical Biology 263: 203–209.
88. Mohabatkar H, Mohammad Beigi M, Esmaeili A (2011) Prediction of
GABA(A) receptor proteins using the concept of Chou’s pseudo-amino acid
composition and support vector machine. Journal of Theoretical Biology 281:
18–23.
89. Lin J, Wang Y (2011) Using a novel AdaBoost algorithm and Chou’s pseudo
amino acid composition for predicting protein subcellular localization. Protein
& Peptide Letters 18: 1219–1225.
90. Gu Q, Ding YS, Zhang TL (2010) Prediction of G-Protein-Coupled Receptor
Classes in Low Homology Using Chou’s Pseudo Amino Acid Composition with
Approximate Entropy and Hydrophobicity Patterns. Protein & Peptide Letters
17: 559–567.
91. Xiao X, Wang P, Chou KC (2012) iNR-PhysChem: A Sequence-Based
Predictor for Identifying Nuclear Receptors and Their Subfamilies via Physical-
Chemical Property Matrix. PLoS ONE 7: e30869.
92. Li YX, Shao YH, Jing L, Deng NY (2011) An efficient support vector machine
approach for identifying protein s-nitrosylation sites. Protein and Peptide
Letters 18: 573–587.
93. Qiu JD, Huang JH, Shi SP, Liang RP (2010) Using the concept of Chou’s
pseudo amino acid composition to predict enzyme family classes: an approach
with support vector machine based on discrete wavelet transform. Protein &
Peptide Letters 17: 715–722.
94. Zhao W, Wang X, Deng R, Wang J, Zhou H (2011) Discrimination of
Thermostable and Thermophilic Lipases using Support Vector Machines.
Protein & Peptide Letters 18: 707–717.
95. Huang T, Chen L, Cai YD, Chou KC (2011) Classification and analysis of
regulatory pathways using graph property, biochemical and physicochemical
property, and functional property. PLoS ONE 6: e25297.
96. He Z, Zhang J, Shi XH, Hu LL, Kong X, et al. (2010) Predicting drug-target
interaction networks based on functional groups and biological features. PLoS
ONE 5: e9603.
97. Li BQ, Hu LL, Niu S, Cai YD, Chou KC (2012) Predict and analyze S-
nitrosylation modification sites with the mRMR and IFS approaches.
J Proteomics 75: 1654–1665.
98. Roy S, Martinez D, Platero H, Lane T, Werner-Washburne M (2009)
Exploiting amino acid composition for predicting protein-protein interactions.
PLoS One 4: e7813.
99. Moses AM, Durbin R (2009) Inferring selection on amino acid preference in
protein domains. Mol Biol Evol 26: 527–536.
100. Angov E, Hillier CJ, Kincaid RL, Lyon JA (2008) Heterologous protein
expression is enhanced by harmonizing the codon usage frequencies of the
target gene with those of the expression host. PLoS One 3: e2189.
101. Goldenberg NM, Steinberg BE (2010) Surface charge: a key determinant of
protein localization and function. Cancer Res 70: 1277–1280.
102. Mbamala EC, Ben-Shaul A, May S (2005) Domain formation induced by the
adsorption of charged proteins on mixed lipid membranes. Biophys J 88: 1702–
1714.
103. Gong S, Park C, Choi H, Ko J, Jang I, et al. (2005) A protein domain
interaction interface database: InterPare. BMC Bioinformatics 6: 207.
104. Li YD, Zhou Z, Lv LX, Hou XP, Li YQ (2009) New approach to achieve high-
level secretory expression of heterologous proteins by using Tat signal peptide.
Protein & Peptide Letters 16: 706–710.
105. Reynolds SM, Kall L, Riffle ME, Bilmes JA, Noble WS (2008) Transmembrane
topology and signal peptide prediction using dynamic bayesian networks. PLoS
Comput Biol 4: e1000213.
106. Saffari B, Mohabatkar H, Mohsenzadeh S (2008) T and B-cell Epitopes
Prediction of Iranian Saffron (Crocus sativus) Profilin by Bioinformatics Tools.
Protein Pept Lett 15: 280–285.
107. Chen J, Liu H, Yang J, Chou KC (2007) Prediction of linear B-cell epitopes
using amino acid pair antigenicity scale. Amino Acids 33: 423–428.
108. Chou KC (1993) A vectorized sequence-coupling model for predicting HIV
protease cleavage sites in proteins. Journal of Biological Chemistry 268: 16938–
16948.
109. Poorman RA, Tomasselli AG, Heinrikson RL, Kezdy FJ (1991) A cumulative
specificity model for proteases from human immunodeficiency virus types 1 and
2, inferred from statistical analysis of an extended substrate data base. Journal
of Biological Chemistry 266: 14554–14561.
110. Chou KC (1996) Review: Prediction of HIV protease cleavage sites in proteins.
Analytical Biochemistry 233: 1–14.
111. Shen HB, Chou KC (2008) HIVcleave: a web-server for predicting HIV
protease cleavage sites in proteins. Analytical Biochemistry 375: 388–390.
112. Chou KC (1995) A sequence-coupled vector-projection model for predicting
the specificity of GalNAc-transferase. Protein Science 4: 1365–1383.
113. Elhammer AP, Poorman RA, Brown E, Maggiora LL, Hoogerheide JG, et al.
(1993) The specificity of UDP-GalNAc:polypeptide N-acetylgalactosaminyl-
transferase as inferred from a database of in vivo substrates and from the in
vitro glycosylation of proteins and peptides. Journal of Biological Chemistry
268: 10029–10038.
Protein Domain Prediction
PLoS ONE | www.plosone.org 14 June 2012 | Volume 7 | Issue 6 | e39308