Tao Liu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Tao Liu has a total of 80 co-authors. Here are the top co-authors by number of shared publications:
- Shaojun Liu
Institute of Basic Medical Sciences
6 shared publications Haiping Que
Institute of Basic Medical Sciences, Beijing, People’s Republic of China
5 shared publicationsZhenlian Ma
Institute of Basic Medical Sciences, 27 Taiping Road, Beijing 100850, People's Republic of China
4 shared publicationsShuqian Jing
State Key Laboratory of Proteomics and, Department of Neurobiology, Institute of Basic Medical Sciences, Beijing, P. R. China
3 shared publications- Wuju Li
Beijing Institute of Basic Medical Sciences
3 shared publications Xin Li
Institute of Basic Medical Sciences, Beijing, People’s Republic of China
3 shared publications
Latest Publications
- Open Access
- 17 April 2020
- Published by Springer Nature in BioData Mining
- Vol. 13 (1) , 1-13
Deep learning methods improve linear B-cell epitope prediction
Background: B-cell epitopes play important roles in vaccine design, clinical diagnosis, and antibody production. Although some models have been developed to predict linear or conformational B-cell epitopes, their performance is still unsatisfactory. Hundreds of thousands of linear B-cell epitope data have accumulated in the Immune Epitope Database (IEDB). These data can be explored using the deep learning methods, in order to create better predictive models for linear B-cell epitopes. Results: After data cleaning, we obtained 240,563 peptide samples with experimental evidence from the IEDB database, including 25,884 linear B-cell epitopes and 214,679 non-epitopes. Based on the peptide center, we adapted each peptide to the same length by trimming or extending. A random portion of the data, with the same amount of epitopes and non-epitopes, were set aside as test dataset. Then a same number of epitopes and non-epitopes were randomly selected from the remaining data to build a classifier with the feedforward deep neural network. We built eleven classifiers to form an ensemble prediction model. The model will report a peptide as an epitope if it was classified as epitope by all eleven classifiers. Then we used the test data set to evaluate the performance of the model using the area value under the receiver operating characteristic (ROC) curve (AUC) as an indicator. We established 40 models to predict linear B-cell epitopes of length from 11 to 50 separately, and found that the AUC value increased with the length and tended to be stable when the length was 38. Repeated results showed that the models constructed by this method were robust. Tested on our and two public test datasets, our models outperformed current major models available. Conclusions: We applied the feedforward deep neural network to the large amount of linear B-cell epitope data with experimental evidence in the IEDB database, and constructed ensemble prediction models with better performance than the current major models available. We named the models as DLBEpitope and provided web services using the models at http://ccb1.bmi.ac.cn:81/dlbepitope/.
- 17 April 2020
- Published by Springer Nature in BioData Mining
- Vol. 13 (1) , 1-13
Open AccessDeep learning methods improve linear B-cell epitope prediction
Background: B-cell epitopes play important roles in vaccine design, clinical diagnosis, and antibody production. Although some models have been developed to predict linear or conformational B-cell epitopes, their performance is still unsatisfactory. Hundreds of thousands of linear B-cell epitope data have accumulated in the Immune Epitope Database (IEDB). These data can be explored using the deep learning methods, in order to create better predictive models for linear B-cell epitopes. Results: After data cleaning, we obtained 240,563 peptide samples with experimental evidence from the IEDB database, including 25,884 linear B-cell epitopes and 214,679 non-epitopes. Based on the peptide center, we adapted each peptide to the same length by trimming or extending. A random portion of the data, with the same amount of epitopes and non-epitopes, were set aside as test dataset. Then a same number of epitopes and non-epitopes were randomly selected from the remaining data to build a classifier with the feedforward deep neural network. We built eleven classifiers to form an ensemble prediction model. The model will report a peptide as an epitope if it was classified as epitope by all eleven classifiers. Then we used the test data set to evaluate the performance of the model using the area value under the receiver operating characteristic (ROC) curve (AUC) as an indicator. We established 40 models to predict linear B-cell epitopes of length from 11 to 50 separately, and found that the AUC value increased with the length and tended to be stable when the length was 38. Repeated results showed that the models constructed by this method were robust. Tested on our and two public test datasets, our models outperformed current major models available. Conclusions: We applied the feedforward deep neural network to the large amount of linear B-cell epitope data with experimental evidence in the IEDB database, and constructed ensemble prediction models with better performance than the current major models available. We named the models as DLBEpitope and provided web services using the models at http://ccb1.bmi.ac.cn:81/dlbepitope/.
- Open Access
- 3 May 2017
- Published by Springer Nature in BMC Genomics
- Vol. 18 (1) , 1-11
Detecting RNA-RNA interactions in E. coli using a modified CLASH method
Bacterial small regulatory RNAs (sRNAs) play important roles in sensing environment changes through sRNA-target mRNA interactions. However, the current strategy for detecting sRNA-mRNA interactions usually combines bioinformatics prediction and experimental verification, which is hampered by low prediction accuracy and low-throughput. Additionally, among the 4736 sequenced bacterial genomes, only about 2164 sRNAs from 319 strains have been described. Furthermore, target mRNAs of only 157 sRNAs have been uncovered. Obviously, highly efficient methods were required to detect sRNA-mRNA interactions in the sequenced genomes. This study aimed to apply a modified CLASH (cross-linking, ligation and sequencing hybrids) method to detect RNA-RNA interactions in E. coli, a model bacterial organism. Statistically significant interactions were detected in 29 transcript pairs. To the best of our knowledge, 24 pairs were reported for the first time and were novel RNA interactions, including tRNA-tRNA, tRNA-ncRNA (non-coding RNA), tRNA-rRNA, rRNA-mRNA, rRNA-ncRNA, rRNA-rRNA, rRNA-IGT (intergenic transcript), and tRNA-IGT interactions. Discovery of novel RNA-RNA interactions in the present study demonstrates that RNA-RNA interactions might be far more complicated than ever expected. New methods may be required to help discover more novel RNA-RNA interactions. The present work describes a high-throughput protocol not only for discovering new RNA interactions, but also directly obtaining base-pairing sequences, which should be useful in assessing RNA structure and interactions.
- 3 May 2017
- Published by Springer Nature in BMC Genomics
- Vol. 18 (1) , 1-11
Open AccessDetecting RNA-RNA interactions in E. coli using a modified CLASH method
Bacterial small regulatory RNAs (sRNAs) play important roles in sensing environment changes through sRNA-target mRNA interactions. However, the current strategy for detecting sRNA-mRNA interactions usually combines bioinformatics prediction and experimental verification, which is hampered by low prediction accuracy and low-throughput. Additionally, among the 4736 sequenced bacterial genomes, only about 2164 sRNAs from 319 strains have been described. Furthermore, target mRNAs of only 157 sRNAs have been uncovered. Obviously, highly efficient methods were required to detect sRNA-mRNA interactions in the sequenced genomes. This study aimed to apply a modified CLASH (cross-linking, ligation and sequencing hybrids) method to detect RNA-RNA interactions in E. coli, a model bacterial organism. Statistically significant interactions were detected in 29 transcript pairs. To the best of our knowledge, 24 pairs were reported for the first time and were novel RNA interactions, including tRNA-tRNA, tRNA-ncRNA (non-coding RNA), tRNA-rRNA, rRNA-mRNA, rRNA-ncRNA, rRNA-rRNA, rRNA-IGT (intergenic transcript), and tRNA-IGT interactions. Discovery of novel RNA-RNA interactions in the present study demonstrates that RNA-RNA interactions might be far more complicated than ever expected. New methods may be required to help discover more novel RNA-RNA interactions. The present work describes a high-throughput protocol not only for discovering new RNA interactions, but also directly obtaining base-pairing sequences, which should be useful in assessing RNA structure and interactions.
- Open Access
- 25 October 2015
- Published by Oxford University Press (OUP) in Nucleic Acids Research
- Vol. 44 (D1) , D248-D253
sRNATarBase 3.0: an updated database for sRNA-target interactions in bacteria
Bacterial sRNAs are a class of small regulatory RNAs of about 40–500 nt in length; they play multiple biological roles through binding to their target mRNAs or proteins. Therefore, elucidating sRNA targets is very important. However, only targets of a few sRNAs have been described. To facilitate sRNA functional studies such as developing sRNA target prediction models, we updated the sRNATarBase database, which was initially developed in 2010. The new version (recently moved to http://ccb1.bmi.ac.cn/srnatarbase/) contains 771 sRNA-target entries manually collected from 213 papers, and 23 290 and 11 750 predicted targets from sRNATarget and sTarPicker, respectively. Among the 771 entries, 475 and 17 were involved in validated sRNA–mRNA and sRNA–protein interactions, respectively, while 279 had no reported interactions. We also presented detailed information for 316 binding regions of sRNA-target mRNA interactions and related mutation experiments, as well as new features, including NCBI sequence viewer, sRNA regulatory network, target prediction-based GO and pathway annotations, and error report system. The new version provides a comprehensive annotation of validated sRNA-target interactions, and will be a useful resource for bacterial sRNA studies.
- 25 October 2015
- Published by Oxford University Press (OUP) in Nucleic Acids Research
- Vol. 44 (D1) , D248-D253
Open AccesssRNATarBase 3.0: an updated database for sRNA-target interactions in bacteria
Bacterial sRNAs are a class of small regulatory RNAs of about 40–500 nt in length; they play multiple biological roles through binding to their target mRNAs or proteins. Therefore, elucidating sRNA targets is very important. However, only targets of a few sRNAs have been described. To facilitate sRNA functional studies such as developing sRNA target prediction models, we updated the sRNATarBase database, which was initially developed in 2010. The new version (recently moved to http://ccb1.bmi.ac.cn/srnatarbase/) contains 771 sRNA-target entries manually collected from 213 papers, and 23 290 and 11 750 predicted targets from sRNATarget and sTarPicker, respectively. Among the 771 entries, 475 and 17 were involved in validated sRNA–mRNA and sRNA–protein interactions, respectively, while 279 had no reported interactions. We also presented detailed information for 316 binding regions of sRNA-target mRNA interactions and related mutation experiments, as well as new features, including NCBI sequence viewer, sRNA regulatory network, target prediction-based GO and pathway annotations, and error report system. The new version provides a comprehensive annotation of validated sRNA-target interactions, and will be a useful resource for bacterial sRNA studies.
- 8 June 2015
- Published by Wiley in Journal of Cellular Physiology
- Vol. 231 (1) , 172-180
Different Effects of p52SHC1 and p52SHC3 on the Cell Cycle of Neurons and Neural Stem Cells
SHC3 is exclusively expressed in postmitotic neurons, while SHC1 is found in neural stem cells and neural precursor cells but absent in mature neurons. In this study, we discovered that suppression of p52SHC1 expression by RNA interference resulted in proliferation defects in neural stem cells, along with significantly reduced protein levels of cyclin E and cyclin A. At the same time, p52SHC3 RNAi caused cell cycle re‐entry (9.54% in S phase and 5.70% in G2‐M phase) in primary neurons with significantly up‐regulated expression of cyclin D1, cyclin E, cyclin A, CDK2, and phosphorylated CDK2. When p52SHC3 was overexpressed, the cell cycle of neural stem cells was arrested with reduced protein levels of cyclin D1, cyclin E, and cyclin A, while overexpression of p52SHC1 did not result in significant changes in postmitotic neurons. Our results indicate that p52SHC3 plays an important role in maintaining the mitotic quiescence of neurons, while p52SHC1 regulates the proliferation of neural stem cells. J. Cell. Physiol. 230: 172–180, 2016.
- 8 June 2015
- Published by Wiley in Journal of Cellular Physiology
- Vol. 231 (1) , 172-180
Different Effects of p52SHC1 and p52SHC3 on the Cell Cycle of Neurons and Neural Stem Cells
SHC3 is exclusively expressed in postmitotic neurons, while SHC1 is found in neural stem cells and neural precursor cells but absent in mature neurons. In this study, we discovered that suppression of p52SHC1 expression by RNA interference resulted in proliferation defects in neural stem cells, along with significantly reduced protein levels of cyclin E and cyclin A. At the same time, p52SHC3 RNAi caused cell cycle re‐entry (9.54% in S phase and 5.70% in G2‐M phase) in primary neurons with significantly up‐regulated expression of cyclin D1, cyclin E, cyclin A, CDK2, and phosphorylated CDK2. When p52SHC3 was overexpressed, the cell cycle of neural stem cells was arrested with reduced protein levels of cyclin D1, cyclin E, and cyclin A, while overexpression of p52SHC1 did not result in significant changes in postmitotic neurons. Our results indicate that p52SHC3 plays an important role in maintaining the mitotic quiescence of neurons, while p52SHC1 regulates the proliferation of neural stem cells. J. Cell. Physiol. 230: 172–180, 2016.
- 28 November 2013
- Published by Springer Nature in In Vitro Cellular & Developmental Biology - Animal
- Vol. 50 (3) , 183-187
Identification and distribution of rRNH1, a gene upregulated after spinal cord primary neuron injury
Our previous study identified and characterized one differentially expressed gene, Rattus norvegicus ribonuclease/angiogenin inhibitor 1 (rRNH1). Transcriptional activity of lots of genes involves in spinal cord injury (SCI) and regeneration, but the mechanisms remain unknown. In the present study, we analyzed the sequences of rRNH1 and examined the expression pattern of rRNH1 in different adult rat tissues. We found the sequences of rRNH1 show high homology to Homo sapiens ribonuclease/angiogenin inhibitor 1; it encoded a protein of 461 amino acid residues and contained 13 leucine-rich repeat motifs. Using real-time quantitative PCR (q-PCR), rRNH1 mRNA was widely expressed in adult rat tissues, especially in the central nervous system. Moreover, rRNH1 protein was found to be upregulated after SCI. Although the precise function of rRNH1 is unknown, its unique expression pattern and upregulation after SCI suggest that rRNH1 might be involved in the succeeding injury and/or regeneration processes of injured spinal cord. All these data make rRNH1 a new interesting start to study the mechanisms of SCI and neuron regeneration.
- 28 November 2013
- Published by Springer Nature in In Vitro Cellular & Developmental Biology - Animal
- Vol. 50 (3) , 183-187
Identification and distribution of rRNH1, a gene upregulated after spinal cord primary neuron injury
Our previous study identified and characterized one differentially expressed gene, Rattus norvegicus ribonuclease/angiogenin inhibitor 1 (rRNH1). Transcriptional activity of lots of genes involves in spinal cord injury (SCI) and regeneration, but the mechanisms remain unknown. In the present study, we analyzed the sequences of rRNH1 and examined the expression pattern of rRNH1 in different adult rat tissues. We found the sequences of rRNH1 show high homology to Homo sapiens ribonuclease/angiogenin inhibitor 1; it encoded a protein of 461 amino acid residues and contained 13 leucine-rich repeat motifs. Using real-time quantitative PCR (q-PCR), rRNH1 mRNA was widely expressed in adult rat tissues, especially in the central nervous system. Moreover, rRNH1 protein was found to be upregulated after SCI. Although the precise function of rRNH1 is unknown, its unique expression pattern and upregulation after SCI suggest that rRNH1 might be involved in the succeeding injury and/or regeneration processes of injured spinal cord. All these data make rRNH1 a new interesting start to study the mechanisms of SCI and neuron regeneration.
- Open Access
- 9 September 2013
- Published by Springer Nature in Oncogene
- Vol. 33 (30) , 3947-3958
Mutations in 3′-long terminal repeat of HERV-W family in chromosome 7 upregulate syncytin-1 expression in urothelial cell carcinoma of the bladder through interacting with c-Myb
Human endogenous retrovirus (HERV) accounts for ∼ 8% of the human genome. Recent studies have reported that multiple HERV genes and long terminal repeats (LTRs) are involved in human tumorigenesis. Here we demonstrated that HERV-W env (syncytin-1) was overexpressed in 75.6% (62/82) of urothelial cell carcinoma (UCC) tissues of the bladder compared with only 6.1% (5/82) of matched tumor-adjacent tissues (PC and 277A>G) were present in 87.8% (72/82) of UCC tissues. Normal 3′-LTR was found in 12.2% (10/82) of UCC tissues compared with 95.1% (78/82) of matched tumor-adjacent tissues (PC mutation, enhanced the syncytin-1 promoter activity and expression. In-silico analysis, electrophoretic mobility shift assays and chromatin immunoprecipitation assays demonstrated the binding of c-Myb to 3′-LTRs when the mutations occurred. This alternative interaction was found to be dependent on 142T>C mutation. C-Myb activated syncytin-1 promoter activity and expression by binding to mutant 3′-LTRs. Taken together, these data indicate that syncytin-1 overexpression may be an indicator of UCC risk. The 3′-LTR mutations may upregulate syncytin-1 expression, enabling it to participate in UCC tumorigenesis and development by interacting with c-Myb.
- 9 September 2013
- Published by Springer Nature in Oncogene
- Vol. 33 (30) , 3947-3958
Open AccessMutations in 3′-long terminal repeat of HERV-W family in chromosome 7 upregulate syncytin-1 expression in urothelial cell carcinoma of the bladder through interacting with c-Myb
Human endogenous retrovirus (HERV) accounts for ∼ 8% of the human genome. Recent studies have reported that multiple HERV genes and long terminal repeats (LTRs) are involved in human tumorigenesis. Here we demonstrated that HERV-W env (syncytin-1) was overexpressed in 75.6% (62/82) of urothelial cell carcinoma (UCC) tissues of the bladder compared with only 6.1% (5/82) of matched tumor-adjacent tissues (PC and 277A>G) were present in 87.8% (72/82) of UCC tissues. Normal 3′-LTR was found in 12.2% (10/82) of UCC tissues compared with 95.1% (78/82) of matched tumor-adjacent tissues (PC mutation, enhanced the syncytin-1 promoter activity and expression. In-silico analysis, electrophoretic mobility shift assays and chromatin immunoprecipitation assays demonstrated the binding of c-Myb to 3′-LTRs when the mutations occurred. This alternative interaction was found to be dependent on 142T>C mutation. C-Myb activated syncytin-1 promoter activity and expression by binding to mutant 3′-LTRs. Taken together, these data indicate that syncytin-1 overexpression may be an indicator of UCC risk. The 3′-LTR mutations may upregulate syncytin-1 expression, enabling it to participate in UCC tumorigenesis and development by interacting with c-Myb.
- 28 April 2013
- Published by Springer Nature in Cell and tissue research
- Vol. 353 (3) , 381-389
Rnh1 promotes differentiation and myelination via RhoA in oligodendrocytes
Increases in Rattus norvegicus ribonuclease/angiogenin inhibitor 1 (Rnh1) are observed in rat primary neuron injury and/or the regeneration process and in differentiated oligodendrocytes. However, the roles of Rnh1 in the central nervous system are still largely unexplored. RhoA is an important signaling protein that has been implicated in oligodendrocyte differentiation and myelination. We demonstrate enhanced differentiation and myelination of oligodendrocytes mediated by Rnh1 in vitro. We further show that Rnh1 is expressed in oligodendrocyte precursors and oligodendrocytes. Importantly, Rnh1 strongly affects oligodendrocyte differentiation through RhoA-ROCK signaling. Moreover, changes in Rnh1 expression in oligodendrocytes regulates the expression and phosphorylation of Fyn, a regulator of RhoA activity. Finally, Rnh1 promotes myelination in vitro. These results show that Rnh1-mediated RhoA inactivation enhances the differentiation and myelination in oligodendrocytes. Overall, Rnh1 might contribute to oligodendrocyte differentiation and myelination processes in vitro.
- 28 April 2013
- Published by Springer Nature in Cell and tissue research
- Vol. 353 (3) , 381-389
Rnh1 promotes differentiation and myelination via RhoA in oligodendrocytes
Increases in Rattus norvegicus ribonuclease/angiogenin inhibitor 1 (Rnh1) are observed in rat primary neuron injury and/or the regeneration process and in differentiated oligodendrocytes. However, the roles of Rnh1 in the central nervous system are still largely unexplored. RhoA is an important signaling protein that has been implicated in oligodendrocyte differentiation and myelination. We demonstrate enhanced differentiation and myelination of oligodendrocytes mediated by Rnh1 in vitro. We further show that Rnh1 is expressed in oligodendrocyte precursors and oligodendrocytes. Importantly, Rnh1 strongly affects oligodendrocyte differentiation through RhoA-ROCK signaling. Moreover, changes in Rnh1 expression in oligodendrocytes regulates the expression and phosphorylation of Fyn, a regulator of RhoA activity. Finally, Rnh1 promotes myelination in vitro. These results show that Rnh1-mediated RhoA inactivation enhances the differentiation and myelination in oligodendrocytes. Overall, Rnh1 might contribute to oligodendrocyte differentiation and myelination processes in vitro.
- 26 April 2012
- Published by Springer Nature in Molecular Biology Reports
- Vol. 39 (7) , 7665-7672
Cloning and characterization of SCIRR69: a novel transcriptional factor belonging to the CREB/ATF family
- 26 April 2012
- Published by Springer Nature in Molecular Biology Reports
- Vol. 39 (7) , 7665-7672
Cloning and characterization of SCIRR69: a novel transcriptional factor belonging to the CREB/ATF family
- Open Access
- 21 April 2011
- Published by Springer Nature in Science China Life Sciences
- Vol. 54 (4) , 336-341
A full-length 3D structure for MAPK/ERK kinase 2 (MEK2)
As a pivotal signal pathway, the Ras/Raf/MEK/ERK cascade can be activated by multiple extracellular stimuli and can transmit signals to diverse substrates. It remains to be elucidated how so many different signals can be variously transferred by only two MEK molecules (MEK1 and MEK2). Because of technological limitations the complete structures of the MEKs are still unavailable. Here, we report the full-length structure of MEK2 obtained by homology modeling and molecular dynamics simulations. The simulations show that the N-terminal part of MEK2 is highly flexible and this flexibility may enable MEK2 to interact with ERKs and other ligands in diverse manners that correspond to various upstream signals and downstream consequences.
- 21 April 2011
- Published by Springer Nature in Science China Life Sciences
- Vol. 54 (4) , 336-341
Open AccessA full-length 3D structure for MAPK/ERK kinase 2 (MEK2)
As a pivotal signal pathway, the Ras/Raf/MEK/ERK cascade can be activated by multiple extracellular stimuli and can transmit signals to diverse substrates. It remains to be elucidated how so many different signals can be variously transferred by only two MEK molecules (MEK1 and MEK2). Because of technological limitations the complete structures of the MEKs are still unavailable. Here, we report the full-length structure of MEK2 obtained by homology modeling and molecular dynamics simulations. The simulations show that the N-terminal part of MEK2 is highly flexible and this flexibility may enable MEK2 to interact with ERKs and other ligands in diverse manners that correspond to various upstream signals and downstream consequences.