Jinyu Yang

Jinyu Yang

Auto-generated profile, based on publicly available information

Jinyu Yang

Department of Mathematics and Statistics, South Dakota State University, Brookings, SD, USA

Research Focus & Metrics

Academic Career Length

11Years (2016-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
19 November 2020
Verified
Book Chapter
Closed access
PDF file
JYJinyu Yang
Jinyu Yang
WAWeizhi An
Weizhi An
SWSheng Wang
Sheng Wang
XZXinliang Zhu
Xinliang Zhu
Chaochao YanChaochao Yan
Chaochao Yan
JHJunzhou Huang
Junzhou Huang

2 August 2019 in Nucleic Acids Research
Verified
Journal Article
Open access
PDF file
JYJinyu Yang
Jinyu Yang
AMANJUN MA
ANJUN MA
AHAdam D Hoppe
Adam D Hoppe
CWCankun Wang
Cankun Wang
YLYang Li
Yang Li
QMQin Ma
Qin Ma
The identification of transcription factor binding sites and cis-regulatory motifs is a frontier whereupon the rules governing protein–DNA binding are being revealed. Here, we developed a new method (DEep Sequence and Shape mOtif or DESSO) for cis-regulatory motif prediction using deep neural networks and the binomial distribution model. DESSO outperformed existing tools, including DeepBind, in predicting motifs in 690 human ENCODE ChIP-sequencing datasets. Furthermore, the deep-learning framework of DESSO expanded motif discovery beyond the state-of-the-art by allowing the identification of known and new protein–protein–DNA tethering interactions in human transcription factors (TFs). Specifically, 61 putative tethering interactions were identified among the 100 TFs expressed in the K562 cell line. In this work, the power of DESSO was further expanded by integrating the detection of DNA shape features. We found that shape information has strong predictive power for TF–DNA binding and provides new putative shape motif information for human TFs. Thus, DESSO improves in the identification and structural analysis of TF binding sites, by integrating the complexities of DNA binding into a deep-learning framework.

13 April 2017 in Bioinformatics
Verified
Journal Article
Open access
PDF file
JYJinyu Yang
Jinyu Yang
XCXin Chen
Xin Chen
AMAdam McDermaid
Adam McDermaid
QMQin Ma
Qin Ma
Motif identification and analyses are important and have been long-standing computational problems in bioinformatics. Substantial efforts have been made in this field during the past several decades. However, the lack of intuitive and integrative web servers impedes the progress of making effective use of emerging algorithms and tools. Here we present an integrated web server, DMINDA 2.0, which contains: (i) five motif prediction and analyses algorithms, including a phylogenetic footprinting framework; (ii) 2125 species with complete genomes to support the above five functions, covering animals, plants and bacteria and (iii) bacterial regulon prediction and visualization. DMINDA 2.0 is freely available at http://bmbl.sdstate.edu/DMINDA2. Supplementary data are available at Bioinformatics online.

6 October 2016 in Bioinformatics
Verified
Journal Article
Open access
PDF file
YZYu Zhang
Yu Zhang
JXJuan Xie
Juan Xie
JYJinyu Yang
Jinyu Yang
AFAnne Fennell
Anne Fennell
CZChi Zhang
Chi Zhang
QMQin Ma
Qin Ma
Biclustering is widely used to identify co-expressed genes under subsets of all the conditions in a large-scale transcriptomic dataset. The program, QUBIC, is recognized as one of the most efficient and effective biclustering methods for biological data interpretation. However, its availability is limited to a C implementation and to a low-throughput web interface. An R implementation of QUBIC is presented here with two unique features: (i) a 82% average improved efficiency by refactoring and optimizing the source C code of QUBIC; and (ii) a set of comprehensive functions to facilitate biclustering-based biological studies, including the qualitative representation (discretization) of expression data, query-based biclustering, bicluster expanding, biclusters comparison, heatmap visualization of any identified biclusters and co-expression networks elucidation. The package is implemented in R (as of version 3.3) and is available from Bioconductor at the URL: http://bioconductor.org/packages/QUBIC, where installation and usage instructions can be found. Supplementary data are available at Bioinformatics online.

8 March 2017 in Briefings in Bioinformatics
Verified
Journal Article
Closed access
PDF file
BLBingqiang Liu
Bingqiang Liu
JYJinyu Yang
Jinyu Yang
YLYang Li
Yang Li
AMAdam McDermaid
Adam McDermaid
QMQin Ma
Qin Ma
Transcription factors are proteins that bind to specific DNA sequences and play important roles in controlling the expression levels of their target genes. Hence, prediction of transcription factor binding sites (TFBSs) provides a solid foundation for inferring gene regulatory mechanisms and building regulatory networks for a genome. Chromatin immunoprecipitation sequencing (ChIP-seq) technology can generate large-scale experimental data for such protein–DNA interactions, providing an unprecedented opportunity to identify TFBSs (a.k.a. cis-regulatory motifs). The bottleneck, however, is the lack of robust mathematical models, as well as efficient computational methods for TFBS prediction to make effective use of massive ChIP-seq data sets in the public domain. The purpose of this study is to review existing motif-finding methods for ChIP-seq data from an algorithmic perspective and provide new computational insight into this field. The state-of-the-art methods were shown through summarizing eight representative motif-finding algorithms along with corresponding challenges, and introducing some important relative functions according to specific biological demands, including discriminative motif finding and cofactor motifs analysis. Finally, potential directions and plans for ChIP-seq-based motif-finding tools were showcased in support of future algorithm development.

21 September 2020
Verified
Conference Paper
Open access
CYChaochao Yan
Chaochao Yan
SWSheng Wang
Sheng Wang
JYJinyu Yang
Jinyu Yang
TXTingyang Xu
Tingyang Xu
JHJunzhou Huang
Junzhou Huang
Molecule generation is to design new molecules with specific chemical properties and further to optimize the desired chemical properties. Following previous work, we encode molecules into continuous vectors in the latent space and then decode the embedding vectors into molecules under the variational autoencoder (VAE) framework. We investigate the posterior collapse problem of the current widely-used RNN-based VAEs for the molecule sequence generation. For the first time, we point out that the underestimated reconstruction loss of VAEs leads to the posterior collapse, and we also provide both analytical and experimental evidences to support our findings. To fix the problem and avoid the posterior collapse, we propose an effective and efficient solution in this work. Without bells and whistles, our method achieves the state-of-the-art reconstruction accuracy and competitive validity score on the ZINC 250K dataset. When generating 10,000 unique valid molecule sequences from the random prior sampling, it costs the JT-VAE 1450 seconds while our method only needs 9 seconds on a regular desktop machine.