Yuting He - Worcester Polytechnic Institute

Yuting He
38 Bowdoin St, Apt 4, Worcester, MA, 01609; 774-578-5183; [email protected], webpage http://users.wpi.edu/~yhe2/
OBJECTIVES
Full time position in Bioinformatics
EDUCATION
M.S. in Biomedical Engineering, in the field Bioinformatics, GPA 3.76/4.0
08/2012-05/2014
Worcester Polytechnic Institute (WPI), Worcester, MA, US; Advisor: Professor Patrick Flaherty
B.S. Measurement and Control Technology and Instrument, GPA 90/100, top 5%
09/2008-07/2012
Tianjin University (TJU), Tianjin, China
RESEARCH INTEREST
Bioinformatics, Machine learning, Data Mining, Big data analysis, NGS data analysis, Bayesian statistics, Graphical model,
Biomarker/variant detection, Biosignal processing, Deep Learning Artificial Neural Networks
AWARDS AND SCHOLARSHIPS
Research Assistantship, supported by PhRMA Foundation Informatics Grant 2013080079
05/2014-09/2014
Graduate Teaching Assistantship, WPI, MA
08/2012-05/2014
RECOMB 2014 Best Poster Award, conference RECOMB, Pittsburgh, PA
04/2014
National Scholarship, Tianjin University, China
2008-2010
SKILLS
Programming: Python, Matlab, C/ C++, Shell scripting, Makefile, R, HTML, CSS, MySQL
Software: Version control/Git, Microsoft office, Photoshop, Latex, BWA, Samtools, GATK, Picard, IGV, VarScan2, MuTect, Strelka
Operation system: Windows, Linux
PUBLICATION
Yuting He, Patrick Flaherty. RVD2: An ultra-sensitive variant detection model for low-depth targeted next-generation sequencing
data, Bioinformatics, 2014. (In review)
POSTERS
Yuting He, Patrick Flaherty. RVD2: An ultra-sensitive variant detection model for low-depth targeted next-generation sequencing
data, the 22th Annual International Conference on Intelligent Systems for Molecular Biology (ISMB), Boston, MA, July, 2014.
Yuting He, Patrick Flaherty. RVD2: An ultra-sensitive variant detection model for low-depth targeted next-generation sequencing
data, the 18th Annual International Conference on Research in Computational Molecular Biology (RECOMB), Pittsburgh, PA, April,
2014. [Best Poster Award].
RESEARCH EXPERIENCE
• Research assistant, Clinical Genomics lab (http://genomics.wpi.edu), WPI
01/2013-Present
RVD2: Ultra-sensitive variant detection model for NGS data
06/2013-Present
Develop a novel algorithm that uses a hierarchical Bayesian model to estimate allele frequency and call rare variants in
heterogeneous high-throughput next-generation sequencing (NGS) DNA data.
 The algorithm is able to detect rare SNPs at as low as 0.1% minor allele frequency given high depth sequence data. The
algorithm achieved at least 30% improvement on false discovery rate in mutation detection comparing with other algorithms.
 The algorithm has been successfully implemented as a Python module and a Linux Command-line interface using
Metropolis-within-Gibbs sampling approach. A variational approach is under developing for lower time complexity.
Deep learning Neural Networks based vital status prediction system
01/2013-05/2013
 Developed a vital status prediction algorithm for breast cancer patient based on deep learning Neural Networks.
 Processed large scale of gene expression data and clinical data from the Cancer Genome Atlas (TCGA) project.
 Detect biomarkers for breast cancer through gene minimization methods using the gene expression data.
• Documents topic classification using adapted LDA model
11/2013-12/2013, WPI
Adapted documents topic classification Bayesian model Latent Dirichlet Allocation (LDA) with a non-informative prior to
improve the objectiveness of the model and avoid hyperparameters estimation. Adapted model was implemented in C and R.
• Speaker Recognition system using filer bank and Vector Quantization (VQ) method
04/2013-05/2013, WPI
Designed and implemented a system in Matlab that can readily identify whether an intercepted human speech signal has
originated from a target human speaker using Vector Quantization (VQ) method.
• Oscillatory Potentials (OP) OP Extraction from Electroretinograms (ERG) signal
11/2012-12/2012, WPI
Developed high accuracy algorithms to exact OP component from ERG signal. Four different techniques were implemented in
Matlab and showed better or comparable performance to state-of-art algorithms.
EMPLOYMENT & SOCIAL ACTIVITY
Research Assistant/ Teaching Assistant, Worcester Polytechnic Institute, Worcester, MA
05/2014-Present
Co-founder and vice-President, Rubik's Cube Association of Tianjin University, China
08/2009-07/2010
PROFESSIONAL AFFILIATIONS
Member, American Association for the Advancement of Science (AAAS)
06/2013
Member, International Society for Computational Biology (ISCB)
06/2014