Lecturer

Hiroki UEDA

PROFILE

Biography
2000.08 BSc,Mathematics, University of Victoria
2000.10 Inter Quest Co.,Ltd.
2001.08 Softwave Corp.
2002.12 Metropolitan Computer Engineer Association
2003.03 M.S, Graduate School of Engineering, Kanazawa Institute of Technology
2006.09 Resercher,Japan Biological Informatics Consortium
2010.04 Intec Inc.
2013.09 PhD, School of Engineering, The University of Tokyo (UTokyo)
2015.04 Resercher, Fujitsu Limited
2018.03 Lecturer, RCAST, UTokyo

FIELD OF INTEREST

With the development of sequencing technology, electronic data yields in biology have been steadily increasing, and it is already a challenging task to process large volumes of data in a conventional method. In addition, in order to extract knowledge from multi modal big data, (ex. Multi-omics data) it is necessary to incorporate the latest Data Science technology, such as cloud computing and machine learning. Research topics include following.

1. Epitranscriptome analysis

Epitranscriptome is transcriptomics with biochemical modifications of RNA. In previous studies, we have developed a bioinformatics method to comprehensively detect inosine-modified sites in the transcriptome at the base level.

2. Cancer genomics

With using next generation sequencer (NGS), it became feasible to detect cancer somatic mutations comprehensively, and NGS is now used as clinical applications, in additions to a research use. Because the allelic fraction of a mutation depends on the tumor purity, local copy number and clonality, it is sometime difficult to call somatic mutation with high accuracy with different specimen. In previous studies, we developed algorithms to calculate somatic mutations, copy number mutations and tumor rates in cancer cells even under noisy low tumor purity conditions.

3. Bioinformatics data analysis using Data Science

In order to find the biological knowledge from biological big data, it is necessary to aggregate data on a cloud and perform distributed processing. We are developing cloud based NGS analysis pipeline using Hadoop / Spark , popular cloud computing framework, and deep learning library.