We are interested in developing novel machine learning and data mining algorithms to accelerate knowledge discovery in life sciences and engineering fields. Currently, our work focuses on multi-label learning, learning from large-scale data, small sample classification, and dimensionality reduction. We also computationally analyze a variety of biological data (e.g., high-throughput expression data (e.g., microarray), protein interaction data, and protein sequence data) in biological pathway understanding, GWAS studies, cancer biology, and healthcare informatics.

We gratefully acknowledge the support from the following sponsors: NSF, DoD, NIH, HRSA, NASA/EPSCoR, JR & Inez Jay Fund, KTEC, KCALSI, and KU.

• KUPS: The University of Kansas Proteomics Service (KUPS) provides high-quality protein-protein interaction (PPI) datasets for researchers who are interested in eulicidating PPIs with 'in silico' methods.
• AAindex: New version of amino acid index
• DDINet: DDINet provides a network of interacting protein domains, which is modeled by an undirected graph where vertices correspond to Pfam domains, and edges represent interactions inferred using our proposed model.
• CSIDOP: CSIDOP is a new method for protein function assignment based on the shared interacting domain patterns extracted from cross-species protein-protein interaction (PPI) data.
• Microarray: Hidden Markov Model (HMM) based algorithm to detect groups of genes functionally similar to a set of input genes from microarray expression data.
• KU GOAL: University of Kansas Gene Ontology Analysis Layer (KU GOAL) contains multiple tools including GO term navigator, Gene Ontology & Gene Products semantic similarity, and Archive of Statistics on Gene Ontology Database

• SNP detection: Markov Blanket-based method, DASSO-MB (Detection of ASSOciations using Markov Blanket) to detect epistatic interactions in case-control GWAS.
• Gene pathways: Hidden Markov Model (HMM) based algorithm to detect groups of genes functionally similar to a set of input genes from microarray expression data.
• Binding sites: Identifying protein interaction sites without using any structure data. Extracting a wide range of features from protein sequences
• PPI prediction: Predicting protein-protein interactions based on domains with random forests framework
• DataKernel: Providing some MATLAB codes for the implementation of kernel-based distance metric learning.

• TCCLS: IEEE Technical Committee on Computational Life Sciences (TCCLS)
• Science Fair: Langston Hughes Science Fair 2011
• CIKM2012: The 21st ACM International Conference on Information and Knowledge Management 2012