Selected publications in 2019

Predicting drug-disease associations with heterogeneous network embedding.
Yang K, Zhao X, Waxman D, Zhao XM.
Chaos. (2019)

In this paper, we propose a method, namely HED (Heterogeneous network Embedding for Drug-disease association), to predict potential associations between drugs and diseases based on a drug-disease heterogeneous network. Specifically, with the heterogeneous network constructed from known drug-disease associations, HED employs network embedding to characterize drug-disease associations and then trains a classifier to predict novel potential drug-disease associations.

Hierarchical graphical model reveals HFR1 bridging circadian rhythm and flower development in Arabidopsis thaliana.
Duren Z, Wang Y, Wang J, Zhao XM, Lv L, Li X, Liu J, Zhu XG, Chen L, Wang Y.
NPJ Syst Biol Appl. (2019)

Here, we proposed a hierarchical graphical model to estimate TF activity from mRNA expression by building TF complexes with protein cofactors and inferring TF’s downstream regulatory network simultaneously. Then we applied our model on flower development and circadian rhythm processes in Arabidopsis thaliana.

EnImpute: imputing dropout events in single cell RNA sequencing data via ensemble learning.
Zhang XF, Ou-Yang L, Yang S, Zhao XM, Hu X, Yan H.
Bioinformatics. (2019)

Imputation of dropout events that may mislead downstream analyses is a key step in analyzing single-cell RNA-sequencing (scRNA-seq) data. We develop EnImpute, an R package that introduces an ensemble learning method for imputing dropout events in scRNA-seq data. EnImpute combines the results obtained from multiple imputation methods to generate a more accurate result.

Identification of Functional Gene Modules Associated With STAT-Mediated Antiviral Responses to White Spot Syndrome Virus in Shrimp.
Zhu G, Li S, Wu J, Li F, Zhao XM.
Frontiers in Physiology (2019)

In this work, based on the gene expression profiles of shrimp with an injection of WSSV and STAT double strand RNA (dsRNA), we constructed a gene co-expression network for shrimp and identified the gene modules that are possibly responsible for STAT-mediated antiviral responses.

DeepPhos: prediction of protein phosphorylation sites with deep learning.
Luo F, Wang M1, Liu Y, Zhao XM, Li A.
Bioinformatics (2019)

In this study we present DeepPhos, a novel deep learning architecture for prediction of protein phosphorylation. Unlike multi-layer convolutional neural networks, DeepPhos consists of densely connected convolutional neuron network blocks which can capture multiple representations of sequences to make final phosphorylation prediction by intra block concatenation layers and inter block concatenation layers.