Dataset Information

MsiDBN: a method of identifying critical proteins in dynamic PPI networks.

ABSTRACT: Dynamics of protein-protein interactions (PPIs) reveals the recondite principles of biological processes inside a cell. Shown in a wealth of study, just a small group of proteins, rather than the majority, play more essential roles at crucial points of biological processes. This present work focuses on identifying these critical proteins exhibiting dramatic structural changes in dynamic PPI networks. First, a comprehensive way of modeling the dynamic PPIs is presented which simultaneously analyzes the activity of proteins and assembles the dynamic coregulation correlation between proteins at each time point. Second, a novel method is proposed, named msiDBN, which models a common representation of multiple PPI networks using a deep belief network framework and analyzes the reconstruction errors and the variabilities across the time courses in the biological process. Experiments were implemented on data of yeast cell cycles. We evaluated our network construction method by comparing the functional representations of the derived networks with two other traditional construction methods. The ranking results of critical proteins in msiDBN were compared with the results from the baseline methods. The results of comparison showed that msiDBN had better reconstruction rate and identified more proteins of critical value to yeast cell cycle process.

SUBMITTER: Zhang Y

PROVIDER: S-EPMC3996968 | biostudies-literature | 2014

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

msiDBN: a method of identifying critical proteins in dynamic PPI networks.

Zhang Yuan Y Du Nan N Li Kang K Feng Jinchao J Jia Kebin K Zhang Aidong A

BioMed research international 20140402

Dynamics of protein-protein interactions (PPIs) reveals the recondite principles of biological processes inside a cell. Shown in a wealth of study, just a small group of proteins, rather than the majority, play more essential roles at crucial points of biological processes. This present work focuses on identifying these critical proteins exhibiting dramatic structural changes in dynamic PPI networks. First, a comprehensive way of modeling the dynamic PPIs is presented which simultaneously analyz ...[more]

PMID: 24800204

Similar Datasets

Project description:With the development of high throughput technologies, there are more and more protein-protein interaction (PPI) networks available, which provide a need for efficient computational tools for network alignment. Network alignment is widely used to predict functions of certain proteins, identify conserved network modules, and study the evolutionary relationship across species or biological entities. However, network alignment is an NP-complete problem, and previous algorithms are usually slow or less accurate in aligning big networks like human vs. yeast. In this study, we proposed a fast yet accurate algorithm called Network Alignment by Integrating Biological Process (NAIGO). Specifically, we first divided the networks into subnets taking the advantage of known prior knowledge, such as gene ontology. For each subnet pair, we then developed a novel method to align them by considering both protein orthologous information and their local structural information. After that, we expanded the obtained local network alignments in a greedy manner. Taking the aligned pairs as seeds, we formulated the global network alignment problem as an assignment problem based on similarity matrix, which was solved by the Hungarian method. We applied NAIGO to align human and Saccharomyces cerevisiae S288c PPI network and compared the results with other popular methods like IsoRank, GRAAL, SANA, and NABEECO. As a result, our method outperformed the competitors by aligning more orthologous proteins or matched interactions. In addition, we found a few potential functional orthologous proteins such as RRM2B in human and DNA2 in S. cerevisiae S288c, which are related to DNA repair. We also identified a conserved subnet with six orthologous proteins EXO1, MSH3, MSH2, MLH1, MLH3, and MSH6, and six aligned interactions. All these proteins are associated with mismatch repair. Finally, we predicted a few proteins of S. cerevisiae S288c potentially involving in certain biological processes like autophagosome assembly.

Project description:Vegetation state is largely driven by climate and the complexity of involved processes leads to non-linear interactions over multiple time-scales. Recently, the role of temporally lagged dependencies, so-called memory effects, has been emphasized and studied using data-driven methods, relying on a vast amount of Earth observation and climate data. However, the employed models are often not able to represent the highly non-linear processes and do not represent time explicitly. Thus, data-driven study of vegetation dynamics demands new approaches that are able to model complex sequences. The success of Recurrent Neural Networks (RNNs) in other disciplines dealing with sequential data, such as Natural Language Processing, suggests adoption of this method for Earth system sciences. Here, we used a Long Short-Term Memory (LSTM) architecture to fit a global model for Normalized Difference Vegetation Index (NDVI), a proxy for vegetation state, by using climate time-series and static variables representing soil properties and land cover as predictor variables. Furthermore, a set of permutation experiments was performed with the objective to identify memory effects and to better understand the scales on which they act under different environmental conditions. This was done by comparing models that have limited access to temporal context, which was achieved through sequence permutation during model training. We performed a cross-validation with spatio-temporal blocking to deal with the auto-correlation present in the data and to increase the generalizability of the findings. With a full temporal model, global NDVI was predicted with R 2 of 0.943 and RMSE of 0.056. The temporal model explained 14% more variance than the non-memory model on global level. The strongest differences were found in arid and semiarid regions, where the improvement was up to 25%. Our results show that memory effects matter on global scale, with the strongest effects occurring in sub-tropical and transitional water-driven biomes.

Dataset Information

MsiDBN: a method of identifying critical proteins in dynamic PPI networks.

Publications

msiDBN: a method of identifying critical proteins in dynamic PPI networks.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets