Plant Physiology Preview. Published on April 18, 2016, as DOI:10.1104/pp.16.00057
1
Authors:
2
Fangyuan Zhang†,1,2, Shiwei Liu†,1, Ling Li†,1, Kaijing Zuo1, Lingxia Zhao1 and
3
Lida Zhang1
4
1, Department of Plant Science, School of Agriculture and Biology, Shanghai Jiao Tong
5
University, Shanghai 200240, China
6
2, Current Address: Key Laboratory of Eco-environments in Three Gorges Reservoir
7
Region (Ministry of Education), SWU-TAAHC Medicinal Plant Joint R&D Centre,
8
School of Life Sciences, Southwest University, Chongqing 400715, China.
9
10
Running title:
11
Inference of protein interaction network in Arabidopsis
12
13
Journal research area:
14
SYSTEMS AND SYNTHETIC BIOLOGY
15
16
†
17
Fangyuan Zhang, Shiwei Liu & Ling Li
These authors contributed equally to this work.
18
19
Corresponding author:
20
Lida Zhang
21
E-mail address: [email protected]
22
Tel: 86-21-34206144; Fax: 86-21-34206144
23
1
Downloaded from on June 14, 2017 - Published by www.plantphysiol.org
Copyright © 2016 American Society of Plant Biologists. All rights reserved.
Copyright 2016 by the American Society of Plant Biologists
24
AUTHOR CONTRIBUTIONS
25
F.Z., S.L. and L.L. carried out data analyses and the experiments. K.Z. and L.Z
26
contributed experimental materials and advice on the experiments. L.Z. conceived of the
27
project, supervised the research, and wrote the manuscript. All authors approved the
28
manuscript.
29
30
This work was supported by the National Basic Research Program of China (Grant No.
31
2013CB733903-03) and the National Natural Science Foundation of China (Grant No.
32
31371229).
33
34
35
36
37
38
39
40
41
42
43
2
Downloaded from on June 14, 2017 - Published by www.plantphysiol.org
Copyright © 2016 American Society of Plant Biologists. All rights reserved.
44
Genome-wide inference of protein interaction network and its
45
application to the study of crosstalk in Arabidopsis abscisic
46
acid signaling
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
3
Downloaded from on June 14, 2017 - Published by www.plantphysiol.org
Copyright © 2016 American Society of Plant Biologists. All rights reserved.
62
ABSTRACT
63
Protein-protein interactions (PPIs) are essential to almost all cellular processes. To
64
better understand the relationships of proteins in Arabidopsis thaliana, we have
65
developed a genome-wide protein interaction network (AraPPINet) that is inferred from
66
both three-dimensional structures and functional evidence and that encompasses
67
316,747 high-confidence interactions among 12,574 proteins. AraPPINet exhibited high
68
predictive power for discovering protein interactions at a 50% true positive rate and for
69
discriminating positive interactions from similar protein pairs at a 70% true positive rate.
70
Experimental evaluation of a set of predicted PPIs demonstrated the ability of
71
AraPPINet to identify novel protein interactions involved in a specific process at an
72
approximately 100-fold greater accuracy than random protein-protein pairs in a test case
73
of abscisic acid (ABA) signaling. Genetic analysis of an experimentally validated,
74
predicted interaction between ARR1 and PYL1 uncovered crosstalk between ABA and
75
cytokinin signaling in the control of root growth. Therefore we demonstrate the power
76
of AraPPINet (http://netbio.sjtu.edu.cn/arappinet/) as a resource for discovering gene
77
function in converging signaling pathways and complex traits in plant.
78
79
80
81
82
83
84
85
86
87
88
89
90
4
Downloaded from on June 14, 2017 - Published by www.plantphysiol.org
Copyright © 2016 American Society of Plant Biologists. All rights reserved.
91
INTRODUCTION
92
Protein-protein interactions (PPIs) are essential to almost all cellular processes,
93
including DNA replication and transcription, signal transduction, and metabolic cycles.
94
Because proteins function through coordinated interaction, it is important to
95
characterize the nature of these relationships. Deciphering the nature of PPIs not only
96
advances our understanding of gene function at the molecular level but also provides
97
insight into complex cellular processes.
98
The importance of understanding PPIs has prompted the development of various
99
computational approaches, such as gene fusion and Rosetta stone (Marcotte et al., 1999),
100
phylogenetic profiling (Pellegrini et al., 1999), correlated expression of gene pairs
101
(Grigoriev, 2001), gene neighbour (Overbeek et al., 1999), interolog (Matthews et al.,
102
2001), and combining multiple sources of biological data (Rhodes et al., 2005), for
103
uncovering protein interactions over the past decade. Recently, computational methods
104
using structural information for PPI prediction have gained much attention due to the
105
rapid growth of the Protein Data Bank (PDB) (Rose et al., 2013). Protein docking is a
106
promising
107
three-dimensional structural information, but it remains a challenging and
108
computationally demanding task to predict PPIs on a genome-wide scale (Wass et al.,
109
2011). Alternatively, knowledge-based methods that utilize the structural similarity of
110
protein pairs to interface a known protein complex have been used (Aytuna et al., 2005;
111
Zhang et al., 2012; Mosca et al., 2013) The common strategy of these structure-based
112
methods is to find a suitable template complex for the two query protein structures; the
113
prediction is then based on the structural similarity of the two protein models to the
114
template complex. An important advantage of structure-based approaches is their ability
115
to identify the putative interface; namely, they are capable of providing more
116
information than any non-structure based method.
117
Arabidopsis thaliana, a small flowering plant, is widely used as a model organism in
118
plant biology. In recent years, several large-scale experimental PPI studies have begun
119
to unravel the complex cellular networks present in Arabidopsis (Arabidopsis
method
for
discovering
protein
interactions
that
is
5
Downloaded from on June 14, 2017 - Published by www.plantphysiol.org
Copyright © 2016 American Society of Plant Biologists. All rights reserved.
based
on
120
Interactome Mapping Consortium, 2011; Jones et al., 2014). According to an empirical
121
estimation of the size of the protein interactome, experimentally determined interactions
122
represent a small fraction (approximately 6%) of the entire protein interactome, and the
123
relationships among most proteins remain to be discovered (Arabidopsis Interactome
124
Mapping Consortium, 2011). To complement existing experimental resources,
125
genome-wide plant PPI networks have been developed by computational approaches
126
(Cui et al., 2008; Lee et al., 2010; Lin et al., 2011; Wang et al., 2012; Wang et al., 2014).
127
However, these methods attempt to predict protein–protein interactions using only
128
non-structural information.
129
Here, we developed a computational approach combining three-dimensional structure
130
with functional information to construct genome-wide PPI networks in Arabidopsis.
131
Comparison of the predicted interactions with experimentally validated data revealed its
132
power for discovering protein interactions and for discriminating positive interactions
133
from similar protein pairs. Experimental evaluation of the predicted PPIs demonstrated
134
the ability of AraPPINet to discovery novel protein interactions involved in specific
135
processes at 100-fold greater accuracy than screens of random protein-protein pairs for
136
the test case of abscisic acid (ABA) signaling. Using AraPPINet guilt-by-association
137
and genetic analysis, we identified and validated a regulator, ARR1, involved in
138
crosstalk in ABA signaling mediated by PYL1. Our findings suggest that AraPPINet
139
could not only advance our understanding of gene function, but could also provide
140
insight into the molecular basis of the complex interplay between signaling pathways in
141
plants.
142
RESULTS
143
Construction of PPI Networks in Arabidopsis
144
We integrated three-dimensional structure and functional information using a machine
145
learning method for genome-wide PPI prediction in Arabidopsis. The PPI model
146
contained four structural and seven non-structural features. The structural features
147
included structural similarity, structural distance, preserved interface size, and fraction
6
Downloaded from on June 14, 2017 - Published by www.plantphysiol.org
Copyright © 2016 American Society of Plant Biologists. All rights reserved.
148
of the preserved interface that had been derived from structural alignments of protein
149
structure homology models to template complexes. Given a PPI model with multiple
150
values for a structural feature, the data with the highest structural similarity was
151
assigned to this interaction model. This approach resulted in approximately 40 million
152
protein
153
approximately 5.3 billion structural alignments (Supplemental Table S1). The seven
154
non-structural features of the interaction model were three gene ontologies, gene
155
coexpression, interolog, Rosetta stone protein, and phylogenetic profile similarity, of
156
which the proportion ranged from 0.2% to 100% of expectations from all possible
157
protein pairs (Supplemental Table S1). Approximately 343 million protein pairs with
158
available data for these 11 features were then classified using a random forests
159
algorithm for PPI predictions, resulting in an Arabidopsis PPI network (AraPPINet) that
160
contained 316,747 high-confidence interacting pairs among 12,574 (48%) of the 26,207
161
Arabidopsis proteins (Supplemental Fig. S1). Almost a third of the interacting protein
162
pairs predicted by AraPPINet were supported by structural evidence (Supplemental Fig.
163
S2). AraPPINet exhibited small-world properties similar to those of other biological
164
networks (Supplemental Fig. S3A), and high clustering coefficient values indicated that
165
the topological structure of AraPPINet was highly modular (Supplemental Fig. S3B).
166
Network analysis revealed that AraPPINet was composed of 1,005 functional modules
167
clustered by a Markov clustering algorithm with an inflation parameter of 1.8 (Van
168
Dongen, 2008), in which the largest module was preferentially associated with
169
transcriptional regulation (Supplemental Fig. S1).
170
Evaluating the Performance of AraPPINet
171
To compare the accuracy of this combined approach with approaches utilizing either
172
structural or non-structural information, a ten-fold cross-validation was carried out
173
using the reference datasets. This evaluation showed that the true positive rate (TPR) of
174
this method reached 49.8% (Supplemental Fig. S4A), while the false positive rate (FPR)
175
remained at the low level of 0.09% (Supplemental Fig. S4B). Furthermore, receiver
176
operating characteristic (ROC) curves showed that the combined method had a higher
interaction
models
containing
structural
information
extracted
7
Downloaded from on June 14, 2017 - Published by www.plantphysiol.org
Copyright © 2016 American Society of Plant Biologists. All rights reserved.
from
177
area under the curve (AUC) value than methods relying on only structural or
178
non-structural information (Fig. 1A). This approach was comparable in performance to
179
other methods over the entire range of the FPR values (Supplemental Fig. S5), which
180
tended to produce fewer false positives and a lower FPR. This result suggested that the
181
combined approach was more efficient than others at separating positives from a large
182
number of negatives at the genome-wide scale.
183
Although cross-validation indicated that the performance of the combined method was
184
superior to those of other PPI predictive models reliant on only structural or
185
non-structural information, these results could not be applied to estimate the combined
186
model’s ability to discovery novel PPIs in practice. Thus, the performance of AraPPINet
187
was tested on an independent dataset of protein interactions determined by
188
high-throughput experiments from public databases. After discarding 190 positive
189
reference interactions, a total of 11,669 protein interactions were remained as the
190
independent dataset for the performance evaluation. Among these interactions, 1,401
191
PPIs were successfully predicted by this method (Fig. 1B). The evaluation indicated our
192
predictive method was powerful in novel PPI discovery at the genome-wide scale.
193
Next, we used two independent datasets to evaluate the performance of AraPPINet
194
compared to that of three other publicly available PPI prediction methods: AtPID (Cui
195
et al., 2008), AtPIN (Brandão et al., 2009) and PAIR (Lin et al., 2011). Among the
196
11,669 protein interactions from high-throughput experiments, approximately 12.0% of
197
protein interactions could be successfully recognized by AraPPINet, which significantly
198
outperformed the AtPID, AtPIN and PAIR methods (Fig. 1C). As further validation of
199
AraPPINet, we tested its performance on a dataset comprising 4,533 newly reported
200
protein interactions in Arabidopsis after March 2014. Of these protein interactions, 443
201
(9.8%) were predicted by AraPPINet (Fig. 1D), which exhibited better TPR than AtPID,
202
AtPIN, and the most recently developed PPI prediction method, PAIR. Furthermore, we
203
also compared the accuracy of different prediction methods using F-measure. From
204
table 1, apparently AraPPINet showed great improvement over all three published PPI
205
prediction methods based on the F1 score.
8
Downloaded from on June 14, 2017 - Published by www.plantphysiol.org
Copyright © 2016 American Society of Plant Biologists. All rights reserved.
206
It is reasonable to estimate the performance of AraPPINet by its ability to discriminate
207
between interacting and non-interacting pairs of proteins with similar sequences. A total
208
of 1,663 interactions between similar proteins in 56 families were used for a
209
performance evaluation. As shown in Fig. 2A, only 0.3% of the mean TPR was
210
expected by chance alone, and the three PPI prediction methods, AtPIN, AtPID and
211
PAIR, had the mean TPRs ranging from 11.8 % to 32.1 % for the tested dataset.
212
Compared with the performance of the other three methods, AraPPINet showed high
213
predictive power for discovering interactions between protein family members with a
214
mean TPR of 69.5%. In addition, interactions between members of three specific
215
transcription factor families identified by high-throughput experiments were used to
216
evaluate the performance of AraPPINet for individual cases, including 255 interactions
217
among 72 MADS, 25 interactions among 17 bZIP, and 418 interactions among 49
218
ARF/IAA transcription factors (de Folter et al., 2005; Ehlert et al., 2006; Vernoux et al.,
219
2011). As shown in Fig. 2B, AraPPINet had AUC values ranging from 0.65 for the
220
MADS family to 0.77 for the BZIP family. The precision of PPI prediction reached
221
22.1% for MADS and 50.6% for ARF/IAA. A number of the interactions between
222
similar members predicted in AraPPINet overlapped highly with those determined
223
through experimental assays (Fig. 2C), suggesting that AraPPINet is a powerful tool for
224
discriminating positive interactions from similar protein pairs within gene families in
225
Arabidopsis.
226
ABA Signaling Network in AraPPINet
227
The ABA signaling pathway is a gene network that plays important roles in numerous
228
biological processes in plants. Although a linear core ABA signal transduction pathway
229
has been established, the complexity of gene regulation suggests that ABA functions
230
through an intricate network that is interwoven with additional cellular processes. Using
231
the core ABA components as query proteins, including 14 ABA receptors, 9 type 2C
232
protein phosphatases, 10 SNF1-related protein kinases 2 (SnRK2) and 10 bZIP
233
transcription factors, we identified 1,912 proteins in AraPPINet predicted to interact
234
with the 43 core components in ABA signaling pathway. The inferred ABA signaling
9
Downloaded from on June 14, 2017 - Published by www.plantphysiol.org
Copyright © 2016 American Society of Plant Biologists. All rights reserved.
235
network contains 7,272 interactions, of which 197 connections have been
236
experimentally proven (Fig. 3A). Among these node proteins, ABI5, ABI1 and ABI2
237
have the top three connections in the ABA signaling network, consistent with data
238
indicating that these three genes play the most important roles in the signal transduction
239
pathway.
240
The accuracy of the inferred ABA signaling network was validated using two sets of
241
experimentally verified interactions. Among the 31 proteins interacting with the core
242
ABA components determined by high-throughput experiments, 7 (22.6%) proteins were
243
successfully predicted by AraPPINet (Fig. 3B). In addition, among 131 newly reported
244
protein interactions involved in ABA signaling, only two interactions were predicted by
245
the published methods AtPIN, AtPID and PAIR. In contrast, AraPPINet recognized 39
246
(29.8%) novel protein interactions in the dataset, approximately 100-fold more than
247
screens of random protein-protein pairs involved in ABA signaling (Fig. 3C). These
248
results suggest that AraPPINet is very useful for identifying novel genes using known
249
genes in this pathway as baits.
250
The functional representation of proteins in the ABA signaling network was performed
251
using plant gene ontology (GO) slim categories (Ashburner et al., 2005). Candidate
252
genes were significantly enriched in specific biological processes such as cellular
253
response to stimulus, regulation of cellular process, and signal transduction (Fig. 3D).
254
The full functional enrichment data can be found in Supplemental Table S2. It is worth
255
noting that genes involved in ABA signaling were highly enriched in the protein binding
256
GO molecular function category, implying that these proteins preferentially interact
257
with core ABA components to perform their functions. Similarly, functional analysis
258
carried out using the Kyoto Encyclopedia of Genes and Genomes (KEGG) database
259
showed that these proteins are significantly enriched in several specific pathways
260
(Supplemental Table S3). In addition to the expected plant hormone signal transduction
261
pathway, pathways involved in plant-pathogen interaction and plant circadian rhythm
262
were identified (Fig. 3E). In addition, genes interacting with core ABA signaling
263
components were significantly altered in response to ABA treatments (Fig. 3F), and
10
Downloaded from on June 14, 2017 - Published by www.plantphysiol.org
Copyright © 2016 American Society of Plant Biologists. All rights reserved.
264
some of them were induced by other plant hormones (Supplemental Fig. S6), suggesting
265
that these interacting proteins may act as nodes between ABA and other hormone
266
signaling pathways.
267
Validating Novel Interactions in ABA Signaling Network
268
Given that AraPPINet successfully discovered novel PPIs involved in ABA signaling,
269
we evaluated the hypothesized interaction-mediated crosstalk between ABA and other
270
signaling pathways. Among the 1,912 proteins interacting with ABA signaling,
271
twenty-nine proteins predicted to interact with the ABA receptors PYL1 (AT5G46790)
272
or ABI5 (AT2G36270) were selected for the validation of their physical interactions by
273
yeast two-hybrid assay (Supplemental Table S4). As shown in Fig. 4A, AT3G16857,
274
which encodes a regulator of the cytokinin response, was found to interact with PYL1
275
by screening 8 candidate partners. In addition, ABI5, another core component in ABA
276
signaling, was selected as bait to identify new binding partners. We found that four
277
novel proteins (AT5G06960, AT3G22840, AT1G12610 and AT1G13260) were
278
determined to interact with ABI5 by screening 21 candidates (Fig. 4A). AT5G06960,
279
better known as TGA5, is a core component in salicylic acid signaling (Zhang et al.,
280
2003), while AT3G22840 is involved in early light response and seed germination
281
(Harari-Steinberg et al., 2001; Rizza et al., 2011). The proteins AT1G12610 and
282
AT1G13260 (RAV1) belong to the B3-AP2 transcription factor family; the former is
283
involved in gibberellic acid (GA) signaling and response to abiotic stresses (Magome et
284
al., 2008; Kang et al., 2011). These results indicated that the accuracy of AraPPINet is
285
more than 17% in the test case of ABA signaling.
286
A recent study showed that RAV1 co-expression with ABI5 enhanced plant resistance to
287
imposed drought stress (Mittal et al., 2014). Thus, bimolecular fluorescence
288
complementation (BiFC) assays were performed to validate the interaction between the
289
proteins RAV1 and ABI5 in plant cells. Constructs encoding ABI5-YFPN and
290
RAV1-YFPC were co-infiltrated into tobacco leaves, resulting in a visible YFP
291
fluorescence signal in nuclei (Fig. 5B). As a control, empty pEG201YN vector was
292
co-infiltrated with RAV1-YFPC, and no fluorescence signal was observed (Fig. 4B).
11
Downloaded from on June 14, 2017 - Published by www.plantphysiol.org
Copyright © 2016 American Society of Plant Biologists. All rights reserved.
293
These results coupled with the phenotype of transgenic cotton suggested that RAV1
294
physically interacted with ABI5 to increase resistance to drought stress in plants (Mittal
295
et al., 2014).
296
The structural interaction model of ARR1 and PYL1 was produced by superimposing
297
homology models on the template of the contact-dependent growth inhibition
298
toxin/immunity complex from Burkholderia pseudomallei (Fig. 5A). The homology
299
model of ARR1 was structurally similar to chain A of the template complex, while the
300
PYL1 model was structurally aligned to chain B. The interaction model has a high
301
interacting probability score of 0.788, arising from both structural similarity and a
302
conserved interface. Similarly, homology models of ABI5 and its four interacting
303
partners were superimposed with their respective homologous template complexes.
304
These structural interaction models showed a significant level of conservation in the
305
interfaces between ABI5 and the four interacting partners (Fig. 5B–E), which resulted in
306
the interaction of protein pairs with different global structures via similar interface
307
architectures. The structural details of these interactions revealed that conserved
308
interfaces could effectively improve the accuracy of discovering PPIs by computational
309
approaches.
310
Identifying Crosstalk in ABA Signaling
311
The PYL1-interacting partner ARR1, a type-B response regulator, is an important
312
regulator involved in the cytokinin signaling pathway (Sakai et al., 2001; Wang et al.,
313
2011). Based on the physical interaction between ARR1 and PYL1, we hypothesized
314
that ARR1 might be involved in regulating the ABA signaling pathway mediated by
315
PYL1. To validate this hypothesis, the response of the type-B arr1,11,12 triple mutant
316
(CS6986) to ABA was tested by the primary root elongation assay to determine whether
317
the cytokinin signaling core regulator ARR1 was also involved in the regulation of ABA
318
signaling (Cary et al., 1995; Leung et al., 1994). The arr1,11,12 triple mutant and
319
wild-type seedlings were grown on Murashige & Skoog (MS) medium, and they
320
showed similar primary root elongation phenotypes (Fig. 6A, B). Compared with
321
wild-type seedlings, the arr1,11,12 mutant seedlings displayed insensitivity to the
12
Downloaded from on June 14, 2017 - Published by www.plantphysiol.org
Copyright © 2016 American Society of Plant Biologists. All rights reserved.
322
cytokinin-mediated inhibition of primary root growth on MS medium containing 10 µM
323
6-benzyladenine (6-BA) cytokinin (Fig. 6A, B). On MS medium supplemented with 10
324
µM ABA, primary root elongation was inhibited in wild-type seedlings (Fig. 6A, B). In
325
contrast, primary root elongation in arr1,11,12 mutant seedlings was not inhibited by
326
ABA. These data indicated that arr1,11,12 mutant seedlings were insensitive to
327
cytokinin and ABA, suggesting that type-B ARRs, including ARR1, might be involved
328
in ABA signaling regulated by the ABA receptor PYL1 in addition to cytokinin
329
signaling (Fig. 6C).
330
DISCUSSION
331
In this study, we developed a computational approach through combining
332
three-dimensional structures with functional evidence to infer the genome-wide PPI
333
network in Arabidopsis. The power of this method for discovering protein interactions
334
as well as predicting interactions between protein family members was demonstrated by
335
a comparison with other publicly available PPI prediction methods as well as by
336
experimentally determined PPI datasets. The predicted AraPPINet network contains
337
316,747 interactions covering nearly half of proteome (Arabidopsis Genome Initiative,
338
2000), which is in agreement with the estimated size of the protein interactome in
339
Arabidopsis (Arabidopsis Interactome Mapping Consortium, 2011). Furthermore,
340
AraPPINet includes many interactions (85.6% of our predicted PPIs) beyond those
341
found by simple interolog prediction from reference model organisms, which suggested
342
that our predictive method was powerful in novel PPI discovery.
343
Structural information has been successfully used to validate or improve large-scale PPI
344
networks (Prieto et al., 2010; Zhang et al., 2012). Our analysis exhibited that structural
345
evidence could increase the reliability of predicted PPI networks by reducing false
346
positives from the large amount of data regarding non-interacting protein pairs at the
347
genome-wide scale (Supplemental Fig. 4B). It is critical that only very low FPR can
348
produce a small enough number of false positives to be used effectively in practice. In
349
addition to structural evidence, functional clues preferentially contributed to the
350
increased coverage of positive interaction prediction (Supplemental Fig. 4A). These two
13
Downloaded from on June 14, 2017 - Published by www.plantphysiol.org
Copyright © 2016 American Society of Plant Biologists. All rights reserved.
351
factors combined to create balance between accuracy and coverage in PPI prediction
352
using AraPPINet.
353
AraPPINet represents a reliable PPI network and was useful for identifying new
354
proteins involved in specific processes using a set of known genes as baits. For example,
355
using network guilt-by-association, we defined an ABA signaling network
356
encompassing more than 7,000 interactions involving approximately 1,950 proteins in
357
Arabidopsis. This protein interaction map greatly expanded the known ABA signaling
358
network in plants. An evaluation based on two independent datasets showed that nearly
359
30% of the known protein interactions involved in ABA signaling were successfully
360
recognized by AraPPINet, indicating that our computational approach for discovering
361
novel PPIs in ABA signaling is comparable in accuracy to high-throughput experiments
362
(von Mering et al., 2002). Using yeast two-hybrid tests on a set of predicted PPIs linked
363
to the ABA and other signaling pathways, five proteins were validated as newly
364
interacting partners of core components in the ABA signaling pathway. Interestingly,
365
ARR1 was predicted to interact with the ABA receptor PYL1 by AraPPINet and was
366
also detected by experimental screening. By genetic analysis, we validated ARR1
367
involvement in the regulation of cytokinin and ABA signaling through its interaction
368
with PYL1. Our findings suggested that AraPPINet could not only advance our
369
understanding of new gene functions but also provide new insight into the crosstalk
370
between pathways such as ABA signaling with respect to diverse biological processes.
371
AraPPINet represents a step towards the goal of computationally reconstructing reliable
372
PPI networks at a genome-wide scale, and this global protein interaction map could
373
facilitate the understanding of the biological organization of genes for specific functions
374
in plants.
375
METHODS
376
Gold Standard Dataset
377
The experimentally determined PPIs for Arabidopsis were extracted from five databases,
378
BioGRID (Chatr-Aryamontri et al., 2015), IntAct (Orchard et al., 2014), DIP (Salwinski
14
Downloaded from on June 14, 2017 - Published by www.plantphysiol.org
Copyright © 2016 American Society of Plant Biologists. All rights reserved.
379
et al., 2004), MINT (Licata et al., 2012), and BIND (Isserlin et Al., 2011). Proteins
380
without a valid TAIR locus or that were undefined in the Arabidopsis genome were
381
removed, resulting in a total of 17,569 PPIs among 6,725 proteins. The PPIs available in
382
different databases are listed in Supplemental Table S5.
383
A dataset derived from those small-scale or high-throughput experiments with at least
384
two supporting publications was used as a primary positive reference. After removing
385
protein self-interactions or interacting proteins encoded by mitochondria or chloroplast
386
genes, it resulted in 5,080 interacting protein pairs as the gold standard dataset
387
(Supplemental Table S6). Simultaneously, a number of protein pairs were randomly
388
chosen to form the negative reference dataset. The negative dataset created by this
389
method may contain several potentially interacting protein pairs; however, given the
390
empirically estimated ratio of interacting protein pairs of 5 to 10 interactions per 10,000
391
protein pairs in the Arabidopsis genome (Arabidopsis Interactome Mapping Consortium,
392
2011), this level of contamination is likely acceptable. By this way, 508,000 protein
393
pairs not overlapping with 17,569 experimentally determined interactions were
394
randomly generated as the negative reference dataset for training the protein interaction
395
classifier.
396
Collection of Structural Information
397
The structure homology models of Arabidopsis proteins were taken from the ModBase
398
database (Pieper et al., 2014). A protein structure was considered to be reliable if it met
399
at least one of following model evaluation criteria: (1) an E-value < 10-5; (2) an MPQS
400
score (ModPipe quality score) ≥ 0.5; (3) a GA341 score ≥ 0.5; or (4) a z-DOPE score <
401
0. When multiple structures were available for a target protein, we chose the
402
representative model with the highest MPQS score. Based on the above criteria, a total
403
of 17,391 structure homology models were collected for Arabidopsis.
404
A total of 90,424 and 88,999 structures involving multiple organisms were available in
405
the PDB (Protein Data Bank) (Rose et al., 2013) and PISA (Proteins, Interfaces,
406
Structures and Assemblies) databases (Xu et al., 2008), respectively. The chain-chain
407
binary interface and the binding sites of protein complexes were generated in the
15
Downloaded from on June 14, 2017 - Published by www.plantphysiol.org
Copyright © 2016 American Society of Plant Biologists. All rights reserved.
408
PIBASE software package with an interatomic distance cutoff of 6.05 Å (Davis and Sali,
409
2015). A total of 91,652 protein complexes with 368,306 interacting chain pairs among
410
317,112 chains were identified for use as templates for protein interaction predictions.
411
Protein structure homology models were aligned to template complexes using TM-align
412
(Zhang and Skolnick, 2005). Approximately 5.3 billion structural alignments were
413
created between the protein structure homology models and the chains of template
414
complexes. With a normalized TM-score cutoff of 0.4 for structural similarity, a total of
415
50,001,762 structural alignments were generated that involved 16,679 protein structure
416
models.
417
Structural features were derived from the structural alignments of protein structure
418
homology models to template complexes. Because two structural alignments are
419
obtained for each protein pair (i.e., protein structure homology model i is aligned to
420
template chain i’, while protein model j is aligned to chain j’; TM-Scorei is the score
421
between protein structure model i to template chain i’, and TM-Scorej is the score
422
between protein model j to chain j’), structural similarity is calculated as the square root
423
of the product of the TM-Scorei and the TM-Scorej. Similarly, structural distance is
424
calculated as the square root of the product of RMSDi and RMSDj. The preserved
425
interface size is defined as the number of interacting residue pairs in the potential
426
protein-protein model preserved in the template complex. The fraction of the preserved
427
interface is the proportion of the conserved interface of the protein-protein model
428
relative to the corresponding interface in the template complex. When multiple values
429
were available for a protein pair or a structural feature, the data with the highest
430
structural similarity to the protein pair was chosen to create an interaction model.
431
Similarity Analysis of Gene Function
432
The GO data used were gene_ontology.1_2.obo (Ashburner et al., 2005). The functional
433
similarity of a gene pair was defined as log(n/N)/log(2/N), where n is the number of
434
genes in the lowest GO category that contained both genes, and N is the total number of
435
genes annotated for the organism. This formula normalizes the functional similarity
436
range from 0 to 1.
16
Downloaded from on June 14, 2017 - Published by www.plantphysiol.org
Copyright © 2016 American Society of Plant Biologists. All rights reserved.
437
Interolog Analysis
438
Interolog analysis was performed similarly to a strategy proposed by Jonsson and Bates
439
(Jonsson and Bates, 2006). A confidence score S for each PPI prediction was assigned.
440
Given a protein pair A and B, S could be defined as follows:
441
S =  ( ISai × ISbi )
N
i =1
442
where A′i and B′i are the corresponding orthologs of A and B, which show an interaction
443
in one model organism (i.e., the protein pair A′i and B′i is called an interolog of protein
444
pair A and B). ISai is the InParanoid score between A′i and A, while ISbi is the
445
InParanoid score between B′i and B. The orthologs of Arabidopsis proteins in E. coli, S.
446
cerevisae, C. elegans, D. melanogaster, M. musculus and H. sapiens were identified by
447
InParanoid with default settings. N is the total number of interologs of protein pair A
448
and B identified in the six model organisms. The experimentally determined PPI
449
datasets used for interolog analysis were obtained from the BioGRID, IntAct, DIP,
450
MINT and BIND databases (Supplemental Table S7).
451
Coexpression Analysis
452
Arabidopsis GeneChip probe-gene mapping was performed as previously described
453
(Harb et al., 2010). The oligonucleotide sequences of probes were mapped to genes
454
from the TAIR10 database using BLASTN with the E-value < 10-5 (Altschul et al.,
455
1997). If two or more probe sets mapped to a single gene, the expression value for that
456
gene was determined by averaging the signals across the probe sets. In this way, a total
457
of 21,122 genes remained after disregarding probe sets that mapped to more than one
458
gene. A total of 1,388 Affymetrix Arabidopsis arrays were obtained from the TAIR Gene
459
Expression Omnibus (http://www.arabidopsis.org). These slides were normalized using
460
RMAExpress software (Bolstad et al., 2003). The correlations between genes were
461
determined by the Pearson correlation coefficient.
462
Phylogenetic Profile Analysis
463
As previously described by Pellegrini and colleagues (Pellegrini et al., 1999),
17
Downloaded from on June 14, 2017 - Published by www.plantphysiol.org
Copyright © 2016 American Society of Plant Biologists. All rights reserved.
464
phylogenetic profiles for all Arabidopsis protein sequences were constructed using
465
BLASTP searches (E-value < 10-10) against a collection of 30 eukaryotic and 660
466
prokaryotic genomes after applying an evolutionary filter for similar genomes. This
467
calculation across collected genomes yielded a 690-dimensional vector representing the
468
presence or absence of homologues of a query protein in these genomes. The probability
469
of two coevolved proteins is given by P as follows:
470
 M  N − M 
 

x  K − x 

P (x| K, M, N) = 1- 
N
0
 
K 
471
where x is the number of the co-occurrence homologues of proteins A and B in the
472
genomes; K and M are the numbers of homologues of proteins A and B, respectively;
473
and N is the total number of collected genomes. The negative log p-value of
474
phylogenetic profile similarity is used for prediction.
475
Rosetta Stone Analysis
476
Rosetta stone proteins were detected as previously described (Marcotte et al., 1999;
477
Bowers et al., 2004). All protein sequences in the Arabidopsis genome were BLASTPed
478
against the nonredundant (nr) database with an E-value less than 10-10. Two
479
non-homologous proteins were aligned over at least 70% of their sequences to different
480
portions of a third protein, and the third protein was referred to as a candidate Rosetta
481
stone protein.
482
Although proteins containing highly conserved domains may not be fused, they are
483
often linked to each other by the Rosetta stone method. To eliminate these confounding
484
Rosetta stone proteins, the probability that two proteins are linked was computed by
485
chance alone. The probability is given by P as follows:
486
 M  N − M 
 

x −1 
x  K − x 

P (x| K, M, N) = 1- 
N
0
 
K 
487
where x is the number of candidate Rosetta stone proteins; K and M are the numbers of
x −1
18
Downloaded from on June 14, 2017 - Published by www.plantphysiol.org
Copyright © 2016 American Society of Plant Biologists. All rights reserved.
488
homologues of proteins A and B, respectively; and N is the total number of sequences in
489
the nr database. The negative log p-value of the Rosetta stone protein is used for
490
prediction.
491
Random Forest Classifier
492
The positive and negative reference sets with 11 features were used to train random
493
forest classifier in R with the setting 500 trees (Liaw and Wiener, 2002). All protein
494
pairs were then classified by the trained random forest model for PPI prediction. The
495
predicted PPI network was drawn using Cytoscape (Cline et al., 2007).
496
Measurement of Classification Performance
497
The positive and negative reference sets were randomly divided into ten subsets of
498
equal size. Each time, nine subsets were used for classifier training, and the remaining
499
subset was used to test the trained classifier. This procedure was repeated ten times
500
using different subsets for training and testing, and a prediction for each interaction was
501
finally generated. The number of TP (true positive), FP (false positive), TN (true
502
negative) and FN (false negative) predictions were counted respectively. TPR (recall) =
503
TP/(TP﹢FN), FPR = FP/(FP﹢TN), precision = TP/(TP﹢FP), F1 score = 2×precision
504
×recall/( precision + recall) and the area under the curve (AUC) values were calculated
505
against the receiver operating characteristic (ROC) curves.
506
Comparison of AraPPINet with Other Predicted Interactomes
507
The performance of AraPPINet was compared with those of three published PPI
508
prediction methods. The AtPID dataset was retrieved from the AtPID database (version
509
4) (Cui et al., 2008). The AtPIN dataset was downloaded from the AtPIN database
510
(Brandão et al., 2009). The latest PAIR dataset was provided by Dr. Chen (version 3)
511
(Lin et al., 2011). The random PPI networks were generated based on AraPPINet by
512
using an edge-shuffling method, which randomly rewired edges while preserving the
513
original degree distribution of AraPPINet.
514
19
Downloaded from on June 14, 2017 - Published by www.plantphysiol.org
Copyright © 2016 American Society of Plant Biologists. All rights reserved.
515
Functional Enrichment Analysis
516
We used plant GO slim categories to characterize the functions of interacting proteins.
517
The statistical enrichment of proteins in a GO category was obtained by comparing the
518
proteins to those in AraPPINet as well as the Arabidopsis genome using Fisher’s exact
519
test in blast2go (Conesa and Götz, 2008).
520
Similarly, KEGG pathway enrichment analysis of proteins was performed using
521
Fisher’s exact test implemented in a Perl script against a reference dataset of AraPPINet
522
and the Arabidopsis genome.
523
Identification of Hormone-Responsive Genes
524
The identification of hormone-responsive genes was based on the normalized
525
expression profiling data (submission number: ME00333, ME00334, ME00344,
526
ME00337, ME00336, ME00343 and ME00335) of Arabidopsis seedlings from the
527
TAIR Gene Expression Omnibus. The average signal intensities of biological replicates
528
for each sample were used to calculate the fold change of gene expression, and
529
differentially expressed genes were identified as having a minimum fold change of two.
530
Yeast Two-Hybrid Assay
531
Plasmids were constructed using Gateway Cloning Technology (Invitrogen). Insertions
532
of PYL1, ABI5 and the coding sequences of candidate genes were prepared using
533
pENTR/D-TOPO. The bait vector pDEST32 and the prey vector pDEST22 were used
534
for the yeast two-hybrid assay. Sets of bait and prey constructs were co-transformed into
535
the AH109 yeast strain. The transformed yeast cells were selected on synthetic minimal
536
double dropout medium deficient in Trp and Leu. Protein interaction tests were assessed
537
on triple dropout medium deficient in Trp, Leu and His. At least six clones were
538
analysed with three repeats and generated similar results.
539
BiFC Analysis
540
The BiFC vectors pEarleyagte201-YN and pEarleygate202-YC were kindly provided by
541
Professor Steven J. Rothstein for analysis. RAV1 was fused to the C-terminal (amino
20
Downloaded from on June 14, 2017 - Published by www.plantphysiol.org
Copyright © 2016 American Society of Plant Biologists. All rights reserved.
542
acids 175-239) of the YFP protein (YFPC) in the vector pEG202YC, which generated
543
the RAV1-YFPC protein. ABI5 was fused to the N-terminal (amino acids 1–174) of the
544
YFP protein (YFPN) in the vector pEG201YN, which generated the ABI5-YFPN
545
protein. The ABI5-YFPN, RAV1-YFPC and empty pEG201-YFPN constructs were
546
introduced into Agrobacterium tumefaciens strain GV3101. Pairs were co-infiltrated
547
into Nicotiana benthamiana. YFP signal was observed after infiltration from 48 to 60
548
hours by confocal laser scanning microscopy (Leica TCS SP5-II). Each observation was
549
repeated at least three times.
550
Plant Materials and Root Growth Assay
551
Arabidopsis thaliana Columbia (Col-0) was used as the wild-type plant. The T-DNA
552
insertion mutant line arr1,11,12 (CS6986) was kindly provided by Dr. Jiawei Wang. For
553
the root growth assay, seeds of different genotypes were surface-sterilized and then
554
maintained in the dark for 3 days at 4 °C to disrupt dormancy. The seeds were sowed
555
onto MS medium containing 1.5% agar. To investigate the inhibition of root growth by
556
cytokinin and ABA, 5-day-old seedlings were transferred onto plates supplemented with
557
6-BA and ABA (Sigma). The primary root growth of seedlings was measured five days
558
after transfer to the plates.
559
560
561
SUPPLEMENTAL DATA
562
Figure S1. Global view of AraPPINet with the top six assigned modules containing
563
at least 200 node proteins.
564
Figure S2. Coverage of features for interactions in AraPPINet.
565
Figure S3. Topological properties of AraPPINet. A) Degree distribution of the node
566
proteins. B) Average clustering coefficient of proteins with the same degree.
567
Figure S4. Comparisons of performance for methods using different sources of
568
information. A) True positive rates of the methods from three different data sources. B)
569
False positive rate of the methods from three different data sources. Error bars represent
21
Downloaded from on June 14, 2017 - Published by www.plantphysiol.org
Copyright © 2016 American Society of Plant Biologists. All rights reserved.
570
the standard error of the mean.
571
Figure S5. Partial ROC curves for PPIs predicted based on different sources of
572
information.
573
Figure S6. Heatmap of genes within the ABA signaling network in response to
574
plant hormones at different times.
575
Table S1. Features of protein-protein pairs in Arabidopsis.
576
Table S2. Enriched GO functional categories of interacting proteins within the
577
ABA signaling network.
578
Table S3. Enriched KEGG pathways of interacting proteins within the ABA
579
signaling network.
580
Table S4. Predicted PPIs were screened with the yeast two-hybrid assay.
581
Table S5. Presence of PPIs in Arabidopsis from various databases.
582
Table S6. The gold standard PPI data from various databases.
583
Table S7. Availability of PPIs in six model organisms from various databases.
584
585
586
ACKNOWLEDGMENTS
587
We are grateful to Dr. Yi Huang (Oil Crops Research Institute, Chinese Academy of
588
Agricultural Sciences) for helpful discussions and for critical reading of the manuscript.
589
We appreciate the support of the computing facility in the PI cluster at Shanghai Jiao
590
Tong University. We thank Jiawei Wang (Institute of Plant Physiology and Ecology,
591
Chinese Academy of Sciences) for kindly providing arr1,11,12 T-DNA insertion mutant
592
(CS6986) seeds. We also thank Professor Steven J. Rothstein (University of Guelph) for
593
providing the BiFC vectors.
594
595
596
597
22
Downloaded from on June 14, 2017 - Published by www.plantphysiol.org
Copyright © 2016 American Society of Plant Biologists. All rights reserved.
598
TABLES
Table
599 1. Comparison of different prediction approaches with F1 score.
The
600 average true positive of random network is calculated from 10 randomized networks.
Precision
601
= predicted PPIs with experimental evidences/all predicted PPIs, and recall =
predicted
602
PPIs with experimental evidences / all experimentally determined PPIs.
Predicted
PPIs All
All
with experimental predicted
experimentally
Precision Recall
F1 score
evidences
PPIs
determined PPIs
RandNet
348
316,747
21,282
0.001
0.016
0.002
AtPID
386
24,418
21,282
0.016
0.018
0.017
AtPIN
449
90,043
21,282
0.005
0.021
0.008
PAIR
1,995
145,355
21,282
0.014
0.094
0.024
AraPPINet 6,040
316,747
21,282
0.019
0.284
0.036
Method
603
604
605
606
607
608
609
610
611
612
23
Downloaded from on June 14, 2017 - Published by www.plantphysiol.org
Copyright © 2016 American Society of Plant Biologists. All rights reserved.
613
FIGURE LEGENDS
614
Figure 1. Evaluation of AraPPINet performance. A) ROC curves for PPIs inferred
615
from different data sources. AUC values are reported in parentheses. B) Venn diagram
616
of predicted overlapping PPIs with experimentally determined interactions. C)
617
Comparison of AraPPINet with other methods based on the high-throughput PPI dataset.
618
D). Comparison of AraPPINet with other methods based on newly reported interactions.
619
The random PPI networks are generated by randomly rewiring edges while preserving
620
the original degree distribution of AraPPINet. The average TPR is calculated from 10
621
randomized networks.
622
623
Figure 2. Evaluation of AraPPINet for predicting interactions between similar
624
protein pairs. A) Comparison of performance for predicting interactions between
625
similar protein pairs. The boxplots indicate inter-quartile ranges of these data. The bar in
626
each boxplot indicates the median. B) ROC curves for AraPPINet-based predictions for
627
the MADS, BZIP and ARF/IAA families. AUC values are reported in parentheses. C)
628
Venn diagrams of the predicted protein interactions overlapping with the experimentally
629
determined interactions of the MADS, BZIP and ARF/IAA families.
630
631
Figure 3. The ABA signaling network in AraPPINet. A) ABA signaling network
632
inferred from AraPPINet. Node size represents a node degree. Core ABA signaling
633
proteins are represented in red nodes, and their interacting partners are represented in
634
green nodes. The predicted protein interactions are shown in grey lines, and
635
experimentally demonstrated interactions are shown in red lines. B) Comparison of
636
AraPPINet with other methods for discovering PPIs in ABA signaling based on the
637
high-throughput PPI dataset. C) Comparison of AraPPINet with other methods for
638
discovering PPIs in ABA signaling based on newly reported interactions. D) Enriched
639
GO functional categories and E) enriched KEGG pathways of proteins interacting with
640
core ABA signaling proteins. F) Enriched ABA-responsive genes within the ABA
641
signaling network.
24
Downloaded from on June 14, 2017 - Published by www.plantphysiol.org
Copyright © 2016 American Society of Plant Biologists. All rights reserved.
642
643
Figure 4. Yeast two-hybrid and BiFC assays for detecting interacting partners of
644
PYL1 or ABI5. A) Two-hybrid validation in yeast of the interactions between PYL1
645
and ARR1 (AT3G16857), as well as between ABI5 and TGA5 (AT5G06960), ELIP
646
(AT3G22840), RAV1 (AT1G13260) and DDF1 (AT1G12610). BD indicates the fusion
647
protein with the Gal4 BD domain. AD indicates the fusion protein with the Gal4 AD
648
domain. GAD or GBD was used as a negative control. +His indicates synthetic dropout
649
medium deficient in Leu and Trp, -His indicates synthetic dropout medium deficient in
650
Leu, Trp and His. B) BiFC assay for ABI5 and RAV1. 35S:ABI5-YFPN and
651
35S:RAV1-YFPC constructs were co-infiltrated into tobacco leaves. YFP signal
652
intensity was detected from 48 h to 60 h after infiltration.
653
654
Figure 5. Structural models of PPIs. A) Structural interaction model for ARR1 and
655
PYL1. Considering the structural template (PDB structure 4G6V) of the
656
contact-dependent growth inhibition toxin/immunity complex (chain A and B), the
657
homology models of ARR1 and PYL1 were structurally superimposed. B) Structural
658
interaction model for DDF1 and ABI5 based on the template complex (PDB structure
659
1RB6). C) Structural interaction model for RAV1 and ABI5 based on the template
660
complex (PDB structure 1IJ2). D) Structural interaction model for ELIP and ABI5 based
661
on the template complex (PDB structure 2WUK). E) Structural interaction model for
662
TGA5 and ABI5 based on the template complex (PDB structure 2Z5I). The homology
663
models of PYL1 and ABI5 are shown in blue, the models for the interacting protein
664
partners are shown in red, and the template complexes are shown in grey. The conserved
665
interfaces in the van der Waals representation are shown in matching colours.
666
667
Figure 6. arr1,11,12 mutant seedlings are insensitivity to the inhibition of CK and
668
ABA on primary root growth. A) Representative examples of wild-type and
669
arr1,11,12 seedlings on MS medium plates or plates with either 10 µM 6-BA or 10 µM
670
ABA. Five-day-old seedlings grown on MS medium were transferred to MS plates or
671
plates supplemented with either 6-BA or ABA. The pictures were taken five days after
25
Downloaded from on June 14, 2017 - Published by www.plantphysiol.org
Copyright © 2016 American Society of Plant Biologists. All rights reserved.
672
the transfer (scale bar = 10 mm). B) Primary root lengths of wild-type and arr1,11,12
673
seedlings at five days after transfer to MS media plates or plates with either 6-BA or
674
ABA. The data represent the mean values and SE (n > 15). The asterisk indicates that
675
the primary root length of arr1,11,12 is significantly longer than that of wild-type on
676
MS medium plates with either 6-BA or ABA (Student's t-test p < 0.05). C) Schematic
677
representations of the interactions between ABA and cytokinin signaling mediated by
678
ARR1 and PYL1.
679
680
681
682
683
684
685
686
687
688
689
REFERENCES
690
1. Altschul SF, Madden TL, Schäffer AA, Zhang J, Zhang Z, Miller W, Lipman DJ.
691
(1997) Gapped BLAST and PSI-BLAST: a new generation of protein database
692
search programs. Nucleic Acids Res. 25:3389-3402.
693
694
695
696
2. Arabidopsis Genome Initiative. (2000) Analysis of the genome sequence of the
flowering plant Arabidopsis thaliana. Nature. 408:796-815.
3. Arabidopsis Interactome Mapping Consortium. (2011) Evidence for network
evolution in an Arabidopsis interactome map. Science. 333:601-607.
697
4. Ashburner M, Ball CA, Blake JA, Botstein D, Butler H, Cherry JM, Davis AP,
698
Dolinski K, Dwight SS, Eppig JT, Harris MA, Hill DP, Issel-Tarver L, Kasarskis A,
699
Lewis S, Matese JC, Richardson JE, Ringwald M, Rubin GM, Sherlock G. (2005)
26
Downloaded from on June 14, 2017 - Published by www.plantphysiol.org
Copyright © 2016 American Society of Plant Biologists. All rights reserved.
700
Gene ontology: tool for the unification of biology. Nat. Genet. 25:25-29.
701
5. Aytuna AS, Gursoy A, Keskin O. (2005) Prediction of protein-protein interactions
702
by combining structure and sequence conservation in protein interfaces.
703
Bioinformatics. 21:2850-2855.
704
6. Bolstad BM, Irizarry RA, Astrand M, Speed TP. (2003) A comparison of
705
normalization methods for high density oligonucleotide array data based on
706
variance and bias. Bioinformatics. 19:185-193.
707
7. Bowers PM, Pellegrini M, Thompson MJ, Fierro J, Yeates TO, Eisenberg D. (2004)
708
Prolinks: a database of protein functional linkages derived from coevolution.
709
Genome Biol. 5:R35.
710
711
8. Brandão MM, Dantas LL, Silva-Filho MC. (2009) AtPIN: Arabidopsis thaliana
protein interaction network. BMC Bioinformatics. 10:454.
712
9. Cary AJ, Liu W, Howell SH. (1995) Cytokinin action is coupled to ethylene in its
713
effects on the inhibition of root and hypocotyl elongation in Arabidopsis thaliana
714
seedlings. Plant Physiol. 107:1075-1082.
715
10. Chatr-Aryamontri A, Breitkreutz BJ, Oughtred R, Boucher L, Heinicke S, Chen D,
716
Stark C, Breitkreutz A, Kolas N, O'Donnell L, Reguly T, Nixon J, Ramage L,
717
Winter A, Sellam A, Chang C, Hirschman J, Theesfeld C, Rust J, Livstone MS,
718
Dolinski K, Tyers M. (2015) The BioGRID interaction database: 2015 update.
719
Nucleic Acids Res. 43: D470-478.
720
11. Cline MS, Smoot M, Cerami E, Kuchinsky A, Landys N, Workman C, Christmas R,
721
Avila-Campilo I, Creech M, Gross B, Hanspers K, Isserlin R, Kelley R, Killcoyne S,
722
Lotia S, Maere S, Morris J, Ono K, Pavlovic V, Pico AR, Vailaya A, Wang PL,
723
Adler A, Conklin BR, Hood L, Kuiper M, Sander C, Schmulevich I, Schwikowski
724
B, Warner GJ, Ideker T, Bader GD. (2007) Integration of biological networks and
725
gene expression data using Cytoscape. Nat Protoc 2: 2366-2382.
726
727
12. Conesa A, Götz S. (2008) Blast2GO: A comprehensive suite for functional analysis
in plant genomics. Int J Plant Genomics. 2008:619832.
728
13. Cui J, Li P, Li G, Xu F, Zhao C, Li Y, Yang Z, Wang G, Yu Q, Li Y, Shi T. (2008)
729
AtPID: Arabidopsis thaliana protein interactome database--an integrative platform
27
Downloaded from on June 14, 2017 - Published by www.plantphysiol.org
Copyright © 2016 American Society of Plant Biologists. All rights reserved.
730
731
732
for plant systems biology. Nucleic Acids Res. 36:D999-D1008.
14. Davis FP, Sali A. (2015) PIBASE: a comprehensive database of structurally defined
protein interfaces. Bioinformatics. 21:1901-1907.
733
15. de Folter S, Immink RG, Kieffer M, Parenicová L, Henz SR, Weigel D, Busscher M,
734
Kooiker M, Colombo L, Kater MM, Davies B, Angenent GC. (2005)
735
Comprehensive interaction map of the Arabidopsis MADS Box transcription
736
factors. Plant Cell. 17:1424-1433.
737
738
16. Van Dongen S. (2008) Graph clustering via a discrete uncoupling process. SIAM J.
Matrix Aanl. A. 30:121-141.
739
17. Ehlert A, Weltmeier F, Wang X, Mayer CS, Smeekens S, Vicente-Carbajosa J,
740
Dröge-Laser W. (2006) Two-hybrid protein-protein interaction analysis in
741
Arabidopsis protoplasts: establishment of a heterodimerization map of group C and
742
group S bZIP transcription factors. Plant J. 46:890-900.
743
18. Grigoriev A. (2001) A relationship between gene expression and protein
744
interactions on the proteome scale: analysis of the bacteriophage T7 and the yeast
745
Saccharomyces cerevisiae. Nucleic Acids Res. 29:3513-3519.
746
19. Harari-Steinberg O, Ohad I, Chamovitz DA. (2001) Dissection of the light signal
747
transduction pathways regulating the two early light-induced protein genes in
748
Arabidopsis. Plant Physiol. 127:986-997.
749
20. Harb A, Krishnan A, Ambavaram MM, Pereira A. (2010) Molecular and
750
physiological analysis of drought stress in Arabidopsis reveals early responses
751
leading to acclimation in plant growth. Plant Physiol. 154:1254-1271.
752
753
21. Isserlin R, El-Badrawi RA, Bader GD. (2011) The Biomolecular Interaction
Network Database in PSI-MI 2.5. Database. baq037.
754
22. Jones AM, Xuan Y, Xu M, Wang RS, Ho CH, Lalonde S, You CH, Sardi MI, Parsa
755
SA, Smith-Valle E, Su T, Frazer KA, Pilot G, Pratelli R, Grossmann G, Acharya BR,
756
Hu HC, Engineer C, Villiers F, Ju C, Takeda K, Su Z, Dong Q, Assmann SM, Chen
757
J, Kwak JM, Schroeder JI, Albert R, Rhee SY, Frommer WB. (2014) Border
28
Downloaded from on June 14, 2017 - Published by www.plantphysiol.org
Copyright © 2016 American Society of Plant Biologists. All rights reserved.
758
control--a membrane-linked interactome of Arabidopsis. Science. 344:711-716.
759
23. Jonsson PF, Bates PA. (2006) Global topological features of cancer proteins in the
760
human interactome. Bioinformatics. 22:2291-2297.
761
24. Kang HG, Kim J, Kim B, Jeong H, Choi SH, Kim EK, Lee HY, Lim PO. (2011)
762
Overexpression of FTL1/DDF1, an AP2 transcription factor, enhances tolerance to
763
cold, drought, and heat stresses in Arabidopsis thaliana. Plant Sci. 180:634-641.
764
25. Lee K, Thorneycroft D, Achuthan P, Hermjakob H, Ideker T. (2010) Mapping plant
765
interactomes using literature curated and predicted protein-protein interaction data
766
sets. Plant Cell. 22:997-1005.
767
26. Leung J, Bouvier-Durand M, Morris PC, Guerrier D, Chefdor F, Giraudat J. (1994)
768
Arabidopsis ABA response gene ABI1: features of a calcium-modulated protein
769
phosphatase. Science. 264:1448-1452.
770
771
27. Liaw A, Wiener M. (2002) Classification and Regression by randomForest. R News.
2:18-22.
772
28. Licata L, Briganti L, Peluso D, Perfetto L, Iannuccelli M, Galeota E, Sacco F,
773
Palma A, Nardozza AP, Santonico E, Castagnoli L, Cesareni G. (2012) MINT, the
774
molecular interaction database: 2012 update. Nucleic Acids Res. 40:D857-D861.
775
29. Lin M, Zhou X, Shen X, Mao C, Chen X. (2011) The predicted Arabidopsis
776
interactome resource and network topology-based systems biology analyses. Plant
777
Cell. 23:911-922.
778
30. Magome H, Yamaguchi S, Hanada A, Kamiya Y, Oda K. (2008) The DDF1
779
transcriptional activator upregulates expression of a gibberellin-deactivating gene,
780
GA2ox7, under high-salinity stress in Arabidopsis. Plant J. 56:613-626.
781
31. Marcotte EM, Pellegrini M, Ng HL, Rice DW, Yeates TO, Eisenberg D. (1999)
782
Detecting protein function and protein-protein interactions from genome sequences.
783
Science. 285:751-753.
784
32. Matthews LR, Vaglio P, Reboul J, Ge H, Davis BP, Garrels J, Vincent S, Vidal M.
785
(2001) Identification of potential interaction networks using sequence-based
786
searches for conserved protein-protein interactions or "interologs". Genome Res.
29
Downloaded from on June 14, 2017 - Published by www.plantphysiol.org
Copyright © 2016 American Society of Plant Biologists. All rights reserved.
787
11:2120-2126.
788
33. Mittal A, Gampala SS, Ritchie GL, Payton P, Burke JJ, Rock CD. (2014) Related to
789
ABA ‐ Insensitive3 (ABI3)/Viviparous1 and AtABI5 transcription factor
790
coexpression in cotton enhances drought stress adaptation. Plant biotechnol. j. 12:
791
578-589.
792
793
34. Mosca R, Céol A, Aloy P. (2013) Interactome3D: adding structural details to
protein networks. Nat Methods. 10:47-53.
794
35. Orchard S, Ammari M, Aranda B, Breuza L, Briganti L, Broackes-Carter F,
795
Campbell NH, Chavali G, Chen C, del-Toro N, Duesbury M, Dumousseau M,
796
Galeota E, Hinz U, Iannuccelli M, Jagannathan S, Jimenez R, Khadake J, Lagreid A,
797
Licata L, Lovering RC, Meldal B, Melidoni AN, Milagros M, Peluso D, Perfetto L,
798
Porras P, Raghunath A, Ricard-Blum S, Roechert B, Stutz A, Tognolli M, van Roey
799
K, Cesareni G, Hermjakob H. (2014) The MIntAct project--IntAct as a common
800
curation platform for 11 molecular interaction databases. Nucleic Acids Res.
801
42:D358-363.
802
36. Overbeek R, Fonstein M, D'Souza M, Pusch GD, Maltsev N. (1999) The use of
803
gene clusters to infer functional coupling. Proc. Natl. Acad. Sci. USA.
804
96:2896-2901.
805
37. Pellegrini M, Marcotte EM, Thompson MJ, Eisenberg D, Yeates TO. (1999)
806
Assigning protein functions by comparative genome analysis: protein phylogenetic
807
profiles. Proc. Natl. Acad. Sci. USA. 96:4285-4288.
808
38. Pieper U, Webb BM, Dong GQ, Schneidman-Duhovny D, Fan H, Kim SJ, Khuri N,
809
Spill YG, Weinkam P, Hammel M, Tainer JA, Nilges M, Sali A. (2014) ModBase, a
810
database of annotated comparative protein structure models and associated
811
resources. Nucleic Acids Res. 42:D336-346.
812
39. Prieto C, De Las Rivas J. (2010) Structural domain-domain interactions: assessment
813
and comparison with protein-protein interaction data to improve the interactome.
814
Proteins. 78:109-117.
815
40. Rhodes
DR,
Tomlins
SA,
Varambally
S,
Mahavisno
V,
30
Downloaded from on June 14, 2017 - Published by www.plantphysiol.org
Copyright © 2016 American Society of Plant Biologists. All rights reserved.
Barrette
T,
816
Kalyana-Sundaram S, Ghosh D, Pandey A, Chinnaiyan AM. (2005) Probabilistic
817
model of the human protein-protein interaction network. Nat. Biotechnol.
818
23:951-959.
819
41. Rizza A, Boccaccini A, Lopez-Vidriero I, Costantino P, Vittorioso P. (2011)
820
Inactivation of the ELIP1 and ELIP2 genes affects Arabidopsis seed germination.
821
New Phytol. 190:896-905.
822
42. Rose PW, Bi C, Bluhm WF, Christie CH, Dimitropoulos D, Dutta S, Green RK,
823
Goodsell DS, Prlic A, Quesada M, Quinn GB, Ramos AG, Westbrook JD, Young J,
824
Zardecki C, Berman HM, Bourne PE. (2013) The RCSB Protein Data Bank: new
825
resources for research and education. Nucleic Acids Res. 41:D475-D482.
826
43. Sakai H, Honma T, Aoyama T, Sato S, Kato T, Tabata S, Oka A. (2001) ARR1, a
827
transcription factor for genes immediately responsive to cytokinins. Science.
828
294:1519-1521.
829
830
44. Salwinski L, Miller CS, Smith AJ, Pettit FK, Bowie JU, Eisenberg D. (2004) The
Database of Interacting Proteins: 2004 update. Nucleic Acids Res. 32:D449-D451.
831
45. Vernoux T, Brunoud G, Farcot E, Morin V, Van den Daele H, Legrand J, Oliva M,
832
Das P, Larrieu A, Wells D, Guédon Y, Armitage L, Picard F, Guyomarc'h S, Cellier
833
C, Parry G, Koumproglou R, Doonan JH, Estelle M, Godin C, Kepinski S, Bennett
834
M, De Veylder L, Traas J. (2011) The auxin signalling network translates dynamic
835
input into robust patterning at the shoot apex. Mol. Syst. Biol. 7:508.
836
46. Von Mering C, Krause R, Snel B, Cornell M, Oliver SG, Fields S, Bork P. (2002)
837
Comparative assessment of large-scale datasets of protein-protein interactions.
838
Nature. 417:399-403.
839
47. Wang C, Marshall A, Zhang D, Wilson ZA. (2012) ANAP: an integrated knowledge
840
base for Arabidopsis protein interaction network analysis. Plant Physiol.
841
158:1523-1533.
842
48. Wang Y, Li L, Ye T, Zhao S, Liu Z, Feng YQ, Wu Y. (2011) Cytokinin antagonizes
843
ABA suppression to seed germination of Arabidopsis by downregulating ABI5
844
expression. Plant J. 68:249-261.
31
Downloaded from on June 14, 2017 - Published by www.plantphysiol.org
Copyright © 2016 American Society of Plant Biologists. All rights reserved.
845
49. Wang Y, Thilmony R, Zhao Y, Chen G, Gu YQ. (2014) AIM: a comprehensive
846
Arabidopsis interactome module database and related interologs in plants. Database
847
(Oxford). bau117.
848
849
50. Wass MN, Fuentes G, Pons C, Pazos F, Valencia A. (2011) Towards the prediction
of protein interaction partners using physical docking. Mol. Syst. Biol. 7:469.
850
51. Xu Q, Canutescu AA, Wang G, Shapovalov M, Obradovic Z, Dunbrack RL Jr.
851
(2008) Statistical analysis of interface similarity in crystals of homologous proteins.
852
J. Mol. Biol. 381:487-507.
853
52. Zhang QC, Petrey D, Deng L, Qiang L, Shi Y, Thu CA, Bisikirska B, Lefebvre C,
854
Accili D, Hunter T, Maniatis T, Califano A, Honig B. (2012) Structure-based
855
prediction of protein-protein interactions on a genome-wide scale. Nature.
856
490:556-560.
857
858
53. Zhang Y, Skolnick J. (2005) TM-align: a protein structure alignment algorithm
based on the TM-score. Nucleic Acids Res. 33:2302-2309.
859
54. Zhang Y, Tessaro MJ, Lassner M, Li X. (2003) Knockout analysis of Arabidopsis
860
transcription factors TGA2, TGA5, and TGA6 reveals their redundant and essential
861
roles in systemic acquired resistance. Plant Cell. 15:2647-2653.
32
Downloaded from on June 14, 2017 - Published by www.plantphysiol.org
Copyright © 2016 American Society of Plant Biologists. All rights reserved.
Figure 1. Evaluation of AraPPINet performance. A) ROC curves for PPIs inferred
from different data sources. AUC values are reported in parentheses. B) Venn diagram
of predicted overlapping PPIs with experimentally determined interactions. C)
Comparison of AraPPINet with other methods based on the high-throughput PPI data
set. D). Comparison of AraPPINet with other methods based on newly reported
interactions. The random PPI networks are generated by randomly rewiring edges while
preserving the original degree distribution of AraPPINet. The average TPR is calculated
from 10 randomized networks.
Downloaded from on June 14, 2017 - Published by www.plantphysiol.org
Copyright © 2016 American Society of Plant Biologists. All rights reserved.
Figure 2. Evaluation of AraPPINet for predicting interactions between similar
protein pairs. A) Comparison of performance for predicting interactions between
similar protein pairs. The boxplots indicate inter-quartile ranges of these data. The bar in
each boxplot indicates the median. B) ROC curves for AraPPINet-based predictions for
the MADS, BZIP and ARF/IAA families. AUC values are reported in parentheses. C)
Venn diagrams of the predicted protein interactions overlapping with the experimentally
determined interactions of the MADS, BZIP and ARF/IAA families.
Downloaded from on June 14, 2017 - Published by www.plantphysiol.org
Copyright © 2016 American Society of Plant Biologists. All rights reserved.
Figure 3. The ABA signaling network in AraPPINet. A) ABA signaling network
inferred from AraPPINet. Node size represents a node degree. Core ABA signaling
proteins are represented in red nodes, and their interacting partners are represented in
green nodes. The predicted protein interactions are shown in grey lines, and
experimentally demonstrated interactions are shown in red lines. B) Comparison of
AraPPINet with other methods for discovering PPIs in ABA signaling based on the
high-throughput PPI data set. C) Comparison of AraPPINet with other methods for
discovering PPIs in ABA signaling based on newly reported interactions. D) Enriched
GO functional categories and E) enriched KEGG pathways of proteins interacting with
core ABA signaling proteins. F) Enriched ABA-responsive genes within the ABA
signaling network.
Downloaded from on June 14, 2017 - Published by www.plantphysiol.org
Copyright © 2016 American Society of Plant Biologists. All rights reserved.
Figure 4. Yeast two-hybrid and BiFC assays for detecting interacting partners of
PYL1 or ABI5. A) Two-hybrid validation in yeast of the interactions between PYL1
and ARR1 (AT3G16857), as well as between ABI5 and TGA5 (AT5G06960), ELIP
(AT3G22840), RAV1 (AT1G13260) and DDF1 (AT1G12610). BD indicates the fusion
protein with the Gal4 BD domain. AD indicates the fusion protein with the Gal4 AD
domain. GAD or GBD was used as a negative control. +His indicates synthetic dropout
medium deficient in Leu and Trp, -His indicates synthetic dropout medium deficient in
Leu, Trp and His. B) BiFC assay for ABI5 and RAV1. 35S:ABI5-YFPN and
35S:RAV1-YFPC constructs were co-infiltrated into tobacco leaves. YFP signal
intensity was detected from 48 h to 60 h after infiltration.
Downloaded from on June 14, 2017 - Published by www.plantphysiol.org
Copyright © 2016 American Society of Plant Biologists. All rights reserved.
Figure 5. Structural models of PPIs. A) Structural interaction model for ARR1 and
PYL1. Considering the structural template (PDB structure 4G6V) of the
contact-dependent growth inhibition toxin/immunity complex (chain A and B), the
homology models of ARR1 and PYL1 were structurally superimposed. B) Structural
interaction model for DDF1 and ABI5 based on the template complex (PDB structure
1RB6). C) Structural interaction model for RAV1 and ABI5 based on the template
complex (PDB structure 1IJ2). D) Structural interaction model for ELIP and ABI5 based
on the template complex (PDB structure 2WUK). E) Structural interaction model for
TGA5 and ABI5 based on the template complex (PDB structure 2Z5I). The homology
models of PYL1 and ABI5 are shown in blue, the models for the interacting protein
partners are shown in red, and the template complexes are shown in grey. The conserved
interfaces in the van der Waals representation are shown in matching colours.
Downloaded from on June 14, 2017 - Published by www.plantphysiol.org
Copyright © 2016 American Society of Plant Biologists. All rights reserved.
Figure 6. arr1,11,12 mutant seedlings are insensitivity to the inhibition of CK and
ABA on primary root growth. A) Representative examples of wild-type and
arr1,11,12 seedlings on MS medium plates or plates with either 10 µM 6-BA or 10 µM
ABA. Five-day-old seedlings grown on MS medium were transferred to MS plates or
plates supplemented with either 6-BA or ABA. The pictures were taken five days after
the transfer (scale bar = 10 mm). B) Primary root lengths of wild-type and arr1,11,12
seedlings at five days after transfer to MS media plates or plates with either 10 µM
6-BA or 10 µM ABA. The data represent the mean values and SE (n > 15). The asterisk
indicates that the primary root length of arr1,11,12 is significantly longer than that of
wild-type on MS medium plates with either 6-BA or ABA (Student's t-test p < 0.05). C)
Schematic representations of the interactions between ABA and cytokinin signaling
mediated by ARR1 and PYL1.
Downloaded from on June 14, 2017 - Published by www.plantphysiol.org
Copyright © 2016 American Society of Plant Biologists. All rights reserved.
Parsed Citations
Altschul SF, Madden TL, Schäffer AA, Zhang J, Zhang Z, Miller W, Lipman DJ. (1997) Gapped BLAST and PSI-BLAST: a new
generation of protein database search programs. Nucleic Acids Res. 25:3389-3402.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Arabidopsis Genome Initiative. (2000) Analysis of the genome sequence of the flowering plant Arabidopsis thaliana. Nature.
408:796-815.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Arabidopsis Interactome Mapping Consortium. (2011) Evidence for network evolution in an Arabidopsis interactome map. Science.
333:601-607.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Ashburner M, Ball CA, Blake JA, Botstein D, Butler H, Cherry JM, Davis AP, Dolinski K, Dwight SS, Eppig JT, Harris MA, Hill DP,
Issel-Tarver L, Kasarskis A, Lewis S, Matese JC, Richardson JE, Ringwald M, Rubin GM, Sherlock G. (2005) Gene ontology: tool
for the unification of biology. Nat. Genet. 25:25-29.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Aytuna AS, Gursoy A, Keskin O. (2005) Prediction of protein-protein interactions by combining structure and sequence
conservation in protein interfaces. Bioinformatics. 21:2850-2855.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Bolstad BM, Irizarry RA, Astrand M, Speed TP. (2003) A comparison of normalization methods for high density oligonucleotide array
data based on variance and bias. Bioinformatics. 19:185-193.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Bowers PM, Pellegrini M, Thompson MJ, Fierro J, Yeates TO, Eisenberg D. (2004) Prolinks: a database of protein functional
linkages derived from coevolution. Genome Biol. 5:R35.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Brandão MM, Dantas LL, Silva-Filho MC. (2009) AtPIN: Arabidopsis thaliana protein interaction network. BMC Bioinformatics.
10:454.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Cary AJ, Liu W, Howell SH. (1995) Cytokinin action is coupled to ethylene in its effects on the inhibition of root and hypocotyl
elongation in Arabidopsis thaliana seedlings. Plant Physiol. 107:1075-1082.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Chatr-Aryamontri A, Breitkreutz BJ, Oughtred R, Boucher L, Heinicke S, Chen D, Stark C, Breitkreutz A, Kolas N, O'Donnell L,
Reguly T, Nixon J, Ramage L, Winter A, Sellam A, Chang C, Hirschman J, Theesfeld C, Rust J, Livstone MS, Dolinski K, Tyers M.
(2015) The BioGRID interaction database: 2015 update. Nucleic Acids Res. 43: D470-478.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Cline MS, Smoot M, Cerami E, Kuchinsky A, Landys N, Workman C, Christmas R, Avila-Campilo I, Creech M, Gross B, Hanspers K,
Isserlin R, Kelley R, Killcoyne S, Lotia S, Maere S, Morris J, Ono K, Pavlovic V, Pico AR, Vailaya A, Wang PL, Adler A, Conklin BR,
Hood L, Kuiper M, Sander C, Schmulevich I, Schwikowski B, Warner GJ, Ideker T, Bader GD. (2007) Integration of biological
networks and gene expression data using Cytoscape. Nat Protoc 2: 2366-2382.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Conesa A, Götz S. (2008) Blast2GO: A comprehensive suite for functional analysis in plant genomics. Int J Plant Genomics.
2008:619832.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Cui J, Li P, Li G, Xu F, Zhao C, Li Y, Yang Z, Wang G, Yu Q, Li Y, Shi T. (2008) AtPID: Arabidopsis thaliana protein interactome
Downloaded from on June 14, 2017 - Published by www.plantphysiol.org
Copyright © 2016 American Society of Plant Biologists. All rights reserved.
database--an integrative platform for plant systems biology. Nucleic Acids Res. 36:D999-D1008.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Davis FP, Sali A. (2015) PIBASE: a comprehensive database of structurally defined protein interfaces. Bioinformatics. 21:1901-1907.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
de Folter S, Immink RG, Kieffer M, Parenicová L, Henz SR, Weigel D, Busscher M, Kooiker M, Colombo L, Kater MM, Davies B,
Angenent GC. (2005) Comprehensive interaction map of the Arabidopsis MADS Box transcription factors. Plant Cell. 17:1424-1433.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Van Dongen S. (2008) Graph clustering via a discrete uncoupling process. SIAM J. Matrix Aanl. A. 30:121-141.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Ehlert A, Weltmeier F, Wang X, Mayer CS, Smeekens S, Vicente-Carbajosa J, Dröge-Laser W. (2006) Two-hybrid protein-protein
interaction analysis in Arabidopsis protoplasts: establishment of a heterodimerization map of group C and group S bZIP
transcription factors. Plant J. 46:890-900.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Grigoriev A. (2001) A relationship between gene expression and protein interactions on the proteome scale: analysis of the
bacteriophage T7 and the yeast Saccharomyces cerevisiae. Nucleic Acids Res. 29:3513-3519.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Harari-Steinberg O, Ohad I, Chamovitz DA. (2001) Dissection of the light signal transduction pathways regulating the two early
light-induced protein genes in Arabidopsis. Plant Physiol. 127:986-997.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Harb A, Krishnan A, Ambavaram MM, Pereira A. (2010) Molecular and physiological analysis of drought stress in Arabidopsis
reveals early responses leading to acclimation in plant growth. Plant Physiol. 154:1254-1271.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Isserlin R, El-Badrawi RA, Bader GD. (2011) The Biomolecular Interaction Network Database in PSI-MI 2.5. Database. baq037.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Jones AM, Xuan Y, Xu M, Wang RS, Ho CH, Lalonde S, You CH, Sardi MI, Parsa SA, Smith-Valle E, Su T, Frazer KA, Pilot G, Pratelli
R, Grossmann G, Acharya BR, Hu HC, Engineer C, Villiers F, Ju C, Takeda K, Su Z, Dong Q, Assmann SM, Chen J, Kwak JM,
Schroeder JI, Albert R, Rhee SY, Frommer WB. (2014) Border control--a membrane-linked interactome of Arabidopsis. Science.
344:711-716.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Jonsson PF, Bates PA. (2006) Global topological features of cancer proteins in the human interactome. Bioinformatics. 22:22912297.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Kang HG, Kim J, Kim B, Jeong H, Choi SH, Kim EK, Lee HY, Lim PO. (2011) Overexpression of FTL1/DDF1, an AP2 transcription
factor, enhances tolerance to cold, drought, and heat stresses in Arabidopsis thaliana. Plant Sci. 180:634-641.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Lee K, Thorneycroft D, Achuthan P, Hermjakob H, Ideker T. (2010) Mapping plant interactomes using literature curated and
predicted protein-protein interaction data sets. Plant Cell. 22:997-1005.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Leung J, Bouvier-Durand M, Morris PC, Guerrier D, Chefdor F, Giraudat J. (1994) Arabidopsis ABA response gene ABI1: features
of a calcium-modulated protein phosphatase. Science. 264:1448-1452.
Pubmed: Author and Title
Downloaded from on June 14, 2017 - Published by www.plantphysiol.org
Copyright © 2016 American Society of Plant Biologists. All rights reserved.
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Liaw A, Wiener M. (2002) Classification and Regression by randomForest. R News. 2:18-22.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Licata L, Briganti L, Peluso D, Perfetto L, Iannuccelli M, Galeota E, Sacco F, Palma A, Nardozza AP, Santonico E, Castagnoli L,
Cesareni G. (2012) MINT, the molecular interaction database: 2012 update. Nucleic Acids Res. 40:D857-D861.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Lin M, Zhou X, Shen X, Mao C, Chen X. (2011) The predicted Arabidopsis interactome resource and network topology-based
systems biology analyses. Plant Cell. 23:911-922.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Magome H, Yamaguchi S, Hanada A, Kamiya Y, Oda K. (2008) The DDF1 transcriptional activator upregulates expression of a
gibberellin-deactivating gene, GA2ox7, under high-salinity stress in Arabidopsis. Plant J. 56:613-626.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Marcotte EM, Pellegrini M, Ng HL, Rice DW, Yeates TO, Eisenberg D. (1999) Detecting protein function and protein-protein
interactions from genome sequences. Science. 285:751-753.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Matthews LR, Vaglio P, Reboul J, Ge H, Davis BP, Garrels J, Vincent S, Vidal M. (2001) Identification of potential interaction
networks using sequence-based searches for conserved protein-protein interactions or "interologs". Genome Res. 11:2120-2126.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Mittal A, Gampala SS, Ritchie GL, Payton P, Burke JJ, Rock CD. (2014) Related to ABA-Insensitive3 (ABI3)/Viviparous1 and AtABI5
transcription factor coexpression in cotton enhances drought stress adaptation. Plant biotechnol. j. 12: 578-589.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Mosca R, Céol A, Aloy P. (2013) Interactome3D: adding structural details to protein networks. Nat Methods. 10:47-53.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Orchard S, Ammari M, Aranda B, Breuza L, Briganti L, Broackes-Carter F, Campbell NH, Chavali G, Chen C, del-Toro N, Duesbury
M, Dumousseau M, Galeota E, Hinz U, Iannuccelli M, Jagannathan S, Jimenez R, Khadake J, Lagreid A, Licata L, Lovering RC,
Meldal B, Melidoni AN, Milagros M, Peluso D, Perfetto L, Porras P, Raghunath A, Ricard-Blum S, Roechert B, Stutz A, Tognolli M,
van Roey K, Cesareni G, Hermjakob H. (2014) The MIntAct project--IntAct as a common curation platform for 11 molecular
interaction databases. Nucleic Acids Res. 42:D358-363.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Overbeek R, Fonstein M, D'Souza M, Pusch GD, Maltsev N. (1999) The use of gene clusters to infer functional coupling. Proc.
Natl. Acad. Sci. USA. 96:2896-2901.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Pellegrini M, Marcotte EM, Thompson MJ, Eisenberg D, Yeates TO. (1999) Assigning protein functions by comparative genome
analysis: protein phylogenetic profiles. Proc. Natl. Acad. Sci. USA. 96:4285-4288.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Pieper U, Webb BM, Dong GQ, Schneidman-Duhovny D, Fan H, Kim SJ, Khuri N, Spill YG, Weinkam P, Hammel M, Tainer JA, Nilges
M, Sali A. (2014) ModBase, a database of annotated comparative protein structure models and associated resources. Nucleic Acids
Res. 42:D336-346.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Prieto C, De Las Rivas J. (2010) Structural domain-domain interactions: assessment and comparison with protein-protein
interaction data to improve the interactome. Proteins. 78:109-117.
Pubmed: Author and Title
Downloaded from on June 14, 2017 - Published by www.plantphysiol.org
Copyright © 2016 American Society of Plant Biologists. All rights reserved.
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Rhodes DR, Tomlins SA, Varambally S, Mahavisno V, Barrette T, Kalyana-Sundaram S, Ghosh D, Pandey A, Chinnaiyan AM. (2005)
Probabilistic model of the human protein-protein interaction network. Nat. Biotechnol. 23:951-959.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Rizza A, Boccaccini A, Lopez-Vidriero I, Costantino P, Vittorioso P. (2011) Inactivation of the ELIP1 and ELIP2 genes affects
Arabidopsis seed germination. New Phytol. 190:896-905.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Rose PW, Bi C, Bluhm WF, Christie CH, Dimitropoulos D, Dutta S, Green RK, Goodsell DS, Prlic A, Quesada M, Quinn GB, Ramos
AG, Westbrook JD, Young J, Zardecki C, Berman HM, Bourne PE. (2013) The RCSB Protein Data Bank: new resources for research
and education. Nucleic Acids Res. 41:D475-D482.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Sakai H, Honma T, Aoyama T, Sato S, Kato T, Tabata S, Oka A. (2001) ARR1, a transcription factor for genes immediately responsive
to cytokinins. Science. 294:1519-1521.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Salwinski L, Miller CS, Smith AJ, Pettit FK, Bowie JU, Eisenberg D. (2004) The Database of Interacting Proteins: 2004 update.
Nucleic Acids Res. 32:D449-D451.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Vernoux T, Brunoud G, Farcot E, Morin V, Van den Daele H, Legrand J, Oliva M, Das P, Larrieu A, Wells D, Guédon Y, Armitage L,
Picard F, Guyomarc'h S, Cellier C, Parry G, Koumproglou R, Doonan JH, Estelle M, Godin C, Kepinski S, Bennett M, De Veylder L,
Traas J. (2011) The auxin signalling network translates dynamic input into robust patterning at the shoot apex. Mol. Syst. Biol.
7:508.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Von Mering C, Krause R, Snel B, Cornell M, Oliver SG, Fields S, Bork P. (2002) Comparative assessment of large-scale datasets of
protein-protein interactions. Nature. 417:399-403.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Wang C, Marshall A, Zhang D, Wilson ZA. (2012) ANAP: an integrated knowledge base for Arabidopsis protein interaction network
analysis. Plant Physiol. 158:1523-1533.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Wang Y, Li L, Ye T, Zhao S, Liu Z, Feng YQ, Wu Y. (2011) Cytokinin antagonizes ABA suppression to seed germination of
Arabidopsis by downregulating ABI5 expression. Plant J. 68:249-261.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Wang Y, Thilmony R, Zhao Y, Chen G, Gu YQ. (2014) AIM: a comprehensive Arabidopsis interactome module database and related
interologs in plants. Database (Oxford). bau117.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Wass MN, Fuentes G, Pons C, Pazos F, Valencia A. (2011) Towards the prediction of protein interaction partners using physical
docking. Mol. Syst. Biol. 7:469.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Xu Q, Canutescu AA, Wang G, Shapovalov M, Obradovic Z, Dunbrack RL Jr. (2008) Statistical analysis of interface similarity in
crystals of homologous proteins. J. Mol. Biol. 381:487-507.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Zhang QC, Petrey D, Deng L, Qiang L, Shi Y, Thu CA, Bisikirska B, Lefebvre C, Accili D, Hunter T, Maniatis T, Califano A, Honig B.
(2012) Structure-based prediction of protein-protein interactions on a genome-wide scale. Nature. 490:556-560.
Downloaded from on June 14, 2017 - Published by www.plantphysiol.org
Copyright © 2016 American Society of Plant Biologists. All rights reserved.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Zhang Y, Skolnick J. (2005) TM-align: a protein structure alignment algorithm based on the TM-score. Nucleic Acids Res. 33:23022309.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Zhang Y, Tessaro MJ, Lassner M, Li X. (2003) Knockout analysis of Arabidopsis transcription factors TGA2, TGA5, and TGA6
reveals their redundant and essential roles in systemic acquired resistance. Plant Cell. 15:2647-2653.
Pubmed: Author and Title
CrossRef: Author and Title
Google Scholar: Author Only Title Only Author and Title
Downloaded from on June 14, 2017 - Published by www.plantphysiol.org
Copyright © 2016 American Society of Plant Biologists. All rights reserved.