Application of a DNA Chip in Hepatocellular Carcinoma
Team Members: Assoc Prof. Lin Feng
Collaborators: -
Analysis of Gene Expression Profile in Ovarian Cancer by Clustering (Master)
Team Members: Assoc Prof. Lin Feng
Collaborators: -
Machine Learning Algorithms for Gene Expression Analysis (Master)
Team Members: Assoc Prof. Lin Feng
Collaborators: -
In Silicon Modeling of Hormone-regulated Gene Expression Network (PhD)
Team Members: Assoc Prof. Lin Feng
Collaborators: -
Reconstruction of Gene Transcriptional Regulatory Network with Nonparametric Statistical Methods (PhD)
Team Members: Assoc Prof. Lin Feng
Collaborators: -
Visual Computing for Structural & Functional Genomics (PhD)
Team Members: Assoc Prof. Lin Feng
Collaborators: -
HPC for Computational Biology (PhD)
Team Members: Assoc Prof. Lin Feng
Collaborators: -
Creating an In Vivo Navigational Cellular Fluorescence Imaging System with Dynamically Optimized Endomicroscopy
A-Star BMRC Fund: SBIC RP C-010/2006, 2007–2010
PI: A/P Seah Hock Soon, Co-PI: A/P Lin Feng, Collaborator: Asst/P Qian Kemao
Collaboration: National Cancer Centre Singapore, Princeton University
This project is aimed at the development of an innovative embedded computing system to turn the volumetric endomicroscope into an In Vivo navigational cellular imaging system, with real-time feedback and dynamic spatial, contrast and temporal optimization. A unique feature of this project is its close link to clinic research. Non-invasive diagnosis of oral cavity malignancies and bladder cancers has a great impact on medical practice, and it has been a research focus in NCCS. A definitive imaging modality and the corresponding real-time diagnostic system is the key to the problem. Therefore, the novel imaging process developed in the project will be immediately transferred to a state-of-the-art clinic setting to provide accurate, sensitive and non-invasive techniques for early diagnosis of oral cavity and bladder cancers.
Real-time Diagnostic Endomicroscopy
AcRF Fund, RGM 35/06, 2007-2009
PI: A/P Lin Feng, Co-PI: A/P Seah Hock Soon,
This is a collaborative research project focuses on clinic research and exploring various photosensitisers and inducers for diagnosis and treatment of oral and bladder cancers. We develop technologies of real-time 2D/3D medical image processing, feature extraction and visualization. Ratio diagnostic algorithms in spatial & frequency domains should be developed for high sensitivities and specificities of over 90% for distinguishing between benign tissue, dysplasia and cancerous tissue. We are also to explore for an innovative parallel computing architecture and realize real-time image processing and feature analysis with the emerging FPGA technology. It can be integrated into the endoscope imaging system to capture suspicious features for an optimized and focused in vivo imaging process, allowing use of the new photosensitisers and inducers for better contrast and quick clearance.
Collaborative Creation and Application of Interactive Digital Media over the Internet
A-Star SERC Fund, 062-101-0034, 2006-2008
PI: Prof Sun Chengzheng, Co-PIs: A/P Seah Hock Soon, A/P Lin Feng, etc.
The goal of our research is to investigate and develop innovative technologies and software tools for supporting Internet-based real-time collaborative application of interactive digital media, and to significantly increase the productivity and quality of digital media professionals. A core technology to be developed is volume modeling which emphasizes its ability of generating objects that are both geometric and topologic complex. Such collaborative modeling systems will finds its applications in various areas where complex objects need to be designed, such as in biomedical modeling. For example, cross-sectional images are acquired through medical imaging such as X-ray CT. An original voxel-based model can be reconstructed and visualized by volume rendering.
Cellular Fluorescence Imaging with Endomicroscopy
AcRF Fund, RG 169/06, 2007-2010
PI: A/P Seah Hock Soon, Co-Investigators: A/P Lin Feng, Asst/P Qian Kemao
We propose to establish a new clinical concept of 3-dimensional virtual histology through real-time microstructure imaging in patients, and to extend the spatial navigation technology to clinical diagnosis and treatment procedures. The ultimate goal is to create a first-of-the-kind in vivo and in situ navigational cellular imaging system, with real-time feedback and dynamic spatial, contrast and temporal optimization for confocal endomicroscopy.
Dynamically Adaptable Neurocomputer and Its Application to Recognition of Steroid Hormone Response Elements
AcRF Fund, RG 50/06, 2006-2008
PI: A/P Lin Feng, Collaborator: Asst/P Valerie Lin
Collaboration: National Cancer Centre Singapore
Research on steroid hormone regulation of gene expression has a significant impact on biomedicine. Of primary importance in the research is recognition of all the Steroid Hormone Response Elements (HREs) in the DNA sequences of a genome concerned, which unfortunately can hardly be done by the current wet-lab technologies. In Silico modeling and recognition of HREs was suggested, but its application to clinic study was barred by the low sensitivity and low specificity of the previous computing technologies. These technologies are based on statistic models such as Position Weight Matrix and Markov Models of DNA sequences, and implemented by static (compiled) programs. The inflexibility of the models inherited from static programming cannot fit the diverse HRE patterns in nature. To solve this problem, we propose a novel computing technology, termed as Dynamically Adaptable Neurocomputer, or DANC. The concept of DANC arises timely with the recent development of high density, run-time reconfigurable hardware, especially dual-module architectures of Field-Programmable Gate Array (FPGA), as well as the remarkable advancement in Artificial Neural Network (ANN) research. The dual-module FPGA makes it possible to design a neural network with the capability of dynamical adaption of topology, which is the key to accurate modeling and recognition of various sequence structures demonstrated in previously verified HREs. The new DANC computation scheme will create numerous virtual neural connections by means of a limited set of links, at different levels of the digital logic, the arithmetic and the neural structures. Delivered from this project will be a unique neurocomputer with high sensitivity and specificity in identification of HREs, which will effectively help clinic study on hormone regulation.
Development of a Real-time Fluorescence Endoscopy Diagnostic System for the Early Detection of Oral and Bladder Cancers
SingHealth Fund, SHF/FG233P/2005, 2006-2008
Co-PI: A/P Seah Hock Soon, Collaborators: A/P Lin Feng, Asst/P Qian Kemao
Collaboration: National Cancer Centre Singapore
The conventional approach for the diagnosis of oral and bladder cancers is white light endoscopy followed by histopathological examination of biopsy samples. However, oral and bladder tumours are mostly superficial and can be difficult to visualise with white light endoscopy. There is a need to develop a more definitive, non-invasive technique for early diagnosis of oral and bladder cancers. Fluorescence endoscopy is a novel imaging technique that offers a non-invasive means to diagnose and stage cancers without the need to conduct biopsies of suspected lesions. It offers the possibility for staging of cancers without the lengthy process involving histopathological examination of biopsy samples. Fast capturing of suspicious features implied by the fluorescence images provides a means for efficient communication, and an optimised and focused in vivo imaging process. Currently images captured using these imaging techniques have to be processed and analysed "off-line", adding a delay to the diagnosis process. This project aims to develop a rapid, real-time image processing and analysis system to be used with fluorescence endoscopy for early diagnosis and staging of oral and bladder cancers. The ultimate aim is to provide an accurate, sensitive and non-invasive real-time cancer diagnosis and staging system that can be used in an outpatient clinical setting.
Gene Expression Profiling with Oligonucleotide Microarray Technology
NTU Biomedical and Pharmaceutical Engineering Cluster SEED Fund, 2006-2008
PI: A/P Lin Feng, Collaborator: Asst/P Zbynek Bozdeck
Collaboration: Zhejiang University
Microarray technology has been used in gene expression profiling as well as for diagnostic purposes. With the rapid expansion of gene databases and the increasing availability of affordable oligonucleotide fabrication, use of long oligonucleotides (50~70nt) has been proven advantageous over other methods as gene expression probes. The main objectives of this project are to develop a new approach combining good specificity with high sensitivity for the formulation of a design directive to develop oligonucleotide selection algorithms; to implement efficient algorithms for probe design with high quality on an high performance computing system that allows real-time probe design; to identify the best sequences for design, by applying cross-referencing sequences across multiple independent databases, verification against the genome sequence of interest, identifying sequences that hit unique splice variants and so on, especially when dealing with complex target mixtures; to fabricate spotted microarrays by robotic printing of the designed oligonucleotides; to develop advanced algorithms for confocal laser scanning and reconstruction of array images; to study supervised and unsupervised clustering methods for sample classification and gene regulation pattern analysis; and to apply the whole system into genomics study.
DNA Chip Design Programme
SCE Research Programme Fund, 2005-2007
PI: A/P Lin Feng, Collaborators: Asst/P Bertil Schmidt, Asst/P Zbynek Bozdeck
We propose a pioneering research program to build our own capability of designing DNA chips, through the development of a web-based server. DNA chip design is analogous to IC design: While the wafer manufacturers are fed with the large-scale integrated circuit designs and produce electronically functioning semiconductor chips, microarry, or DNA chip, makers rely on various genome-based and pharmacogenetic chip designs to fabricate, for example, biologically workable human genome microarrys. The DNA chip technology is empowering the scientific community to understand the fundamental aspects underlining the growth and development of life as well as to explore the genetic causes of anomalies occurring in the functioning of the human body. Capability of providing DNA chip makers and users with professional designs is of great market value.
|
