Seon-Hee Park

Finding the evidence for protein-protein interactions from PubMed abstracts

MOTIVATION Protein-protein interactions play critical roles in biological processes, and many biologists try to find or to predict crucial information concerning these interactions. Before verifying interactions in biological laboratory work, validating them from previous research is necessary. Although many efforts have been made to create databases that store verified information in a structured form, much interaction information still remains as unstructured text. As the amount of new publications has increased rapidly, a large amount of research has sought to extract interactions from the text automatically. However, there remain various difficulties associated with the process of applying automatically generated results into manually annotated databases. For interactions that are not found in manually stored databases, researchers attempt to search for abstracts or full papers. RESULTS As a result of a search for two proteins, PubMed frequently returns hundreds of abstracts. In this paper, a method is introduced that validates protein-protein interactions from PubMed abstracts. A query is generated from two given proteins automatically and abstracts are then collected from PubMed. Following this, target proteins and their synonyms are recognized and their interaction information is extracted from the collection. It was found that 67.37% of the interactions from DIP-PPI corpus were found from the PubMed abstracts and 87.37% of interactions were found from the given full texts. AVAILABILITY Contact authors.

ADL Monitoring System Using FSR Arrays and Optional 3-Axis Accelerometer

This paper deals with Activities of Daily Living (ADL) Monitoring System. The proposed system takes into account deploying in real home. The important issue in deployment is the noninvasiveness. That is, the user should not feel inconvenience. Therefore, our system has been developed by making use of FSR sensors and an optional small body-activity sensor. In particular, FSR sensor is a typical noninvasive sensor since it has a shape of film. In order to make a light-weight monitoring system, we use as small number of sensors as possible. And we adopt rule-based ADL inferring algorithms to avoid inconvenience in collecting training data for supervised learning. For the purpose of improving the accuracy of occupation/usage detection, we make FSR sensors into FSR array sensors. We evaluate the proposed system in laboratory and real home environment.

Detection abnormal pattern in activities of daily living using sequence alignment method

As the aging is rapidly coming, the necessity of cares for old people increases. As requests for care services for patients requiring help of others increases, various systems for care services have been developed. For care services, recently, there have been developed technologies of tracking and monitoring daily activities of a person and recognizing the daily activities of the person by analyzing tracked data. Particularly, there have been developed systems for taking care of a person whose state should be periodically checked, such as patients or old persons. General care systems are good for tracking what activity the person executes, but limited to detecting what a persons state is, such as a normal or an abnormal state. So it is necessary to develop a new computational method to detect abnormal signs in a life pattern via changes of a sequence of activities of daily living.

Extracting protein-protein interactions in biomedical literature using an existing syntactic parser

We are developing an information extraction system for life science literature. We are currently focusing on PubMed abstracts and trying to extract named entities and their relationships, especially protein names and protein-protein interactions. We are adopting methods including natural language processing, machine learning, and text processing. But we are not developing a new tagging or parsing technique. Developing a new tagger or a new parser specialized in life science literature is a very complex job. And it is not easy to get a good result by tuning an existing parser or by training it without a sufficient corpus. These all are another research topics and we are trying to extract information, not to develop something to help the extracting job or else. In this paper, we introduce our method to use an existing full parser without training or tuning. After tagging sentences and extracting proteins, we make sentences simple by substituting some words like named entities, nouns into one word. Then parsing errors are reduced and parsing precision is increased by this sentence simplification. We parse the simplified sentences syntactically with an existing syntactic parser and extract protein-protein interactions from its results. We show the effects of sentence simplification and syntactic parsing.

A Compact Tunable VCSEL and a Built-in Wavelength Meter for a Portable Optical Resonant Reflection Biosensor Reader

This study reports a portable and precision photonic biosensor reader that can measure the concentration of a particular antigen using an optical resonant reflection biosensor (ORRB). To create a compact biosensor reader, a compact tunable vertical-cavity surface-emitting laser (VCSEL) and a compact built-in wavelength meter were manufactured. The wavelength stability and accuracy of the compact built-in wavelength meter were measured to be less than 0.02 nm and 0.06 nm, respectively. The tunable VCSEL emission wavelength was measured with the compact built-in wavelength meter, it has a fast sweep time (~ 10 seconds) and a wide tuning range (> 4 nm) that are sufficient for biosensor applications based on ORRB. The reflection spectrum of a plastic based ORRB chip was measured by the fabricated portable photonic biosensor reader using the VCSEL and wavelength meter. Although the reader is the size of a palmtop device, it could make a precise measurement of the peak wavelength on equal terms with a conventional bulky optical spectrometer.

Sub-ppm level MEMS gas sensor

A sub-ppm level MEMS gas sensor that can be used for the detection of formaldehyde (HCHO) is presented. It is realized by using a zinc oxide (ZnO) thin-film material with a Ni-seed layer as a sensing material and by bulk micromachining technology. To enhance sensitivity of the MEMS gas sensor with Ni-seed layer was embedded with ZnO sensing material and sensing electrodes. As experimental results, the changed sensor resistance ratio for HCHO gas was 9.65 % for 10 ppb, 18.06 % for 100 ppb, and 35.7 % for 1 ppm, respectively. In addition, the minimum detection level of the fabricated MEMS gas sensor was 10 ppb for the HCHO gas. And the measured output voltage was about 0.94 V for 10 ppb HCHO gas concentration. The noise level of the fabricated MEMS gas sensor was about 50 mV. The response and recovery times were 3 and 5 min, respectively. The consumption power of the Pt micro-heater under sensor testing was 184 mW and its operating temperature was .

Surface modification and immunoassys on COC, cross-flow microfluid channels and FRET molecules

In an effort to manufacture a high-performance disposable microfluidic system for applications in medical diagnostics, a series of studies were carried. Surface modification on COC, nanoparticle-based immunoassays on COC, manufacturing cross flow microfluidic system, synthesis of a pair of molecules for fluorescence resonance energy transfer (FRET) and preparation of a FRET nanoparticle are presented in this report. Plastics are advantageous compared with the conventional hard materials such as glass and silicon, in the application of microfluidic devices. However most plastics lack proper functional groups or hydrophilicity, which are essential requirements for microfluidic devices that need various surface chemistry. Among the plastic materials, cyclo-olefin copolymer (COC) has been recently reported to be equipped with more superior physical and optical properties. However since the COC is synthesized by copolymerizing ethylene with norbornene based on Metallocene catalyst, COC is composed only of C and H with no unsaturation. The saturated hydrocarbon polymer provides high optical clarity, which is also the cause for inertness of the COC surface. In an effort to find effective application as a microfluidic device for diagnostic purpose, we have generated an effective functional group (-OH) on the surface of COC by treating the COC with O2 plasma. The surface was characterized by ESCA and the fluorescence microscopy. We further proceeded to carry out sandwich-type immunoassays using fluorescence nanoparticles. A series of nanoparticle-based immunoassays were processed on a COC plate and the results were compared with those on a glass plate. AFP was turned out to be detected as low as a few ng/ml. Microfluidic channels for cross reactions were manufactured. Two types of fluidic channels were manufactured; one for surface modification and the other immunoassays. Also a series of fluorescence molecules were synthesized to choose an optimum pair for FRET. The behavior of fluorescence and FRET of the acceptor and donor dyes were studied. The future research will concern integration of each component for a disposable microfluidic chip.

Semi-Automated Clustering of Gene Expression Data Sets

Clustering, as one of key analysis tools for gene expression data sets, attempts to discover groups of genes having similar expression patterns. In order to get a reasonable biological interpretation, it is desirable that a clustering result be accurate enough. However, conventional clustering methods do not always meet this demand since they require the exact tuning of input parameters and cluster centers for an acceptable quality of result. Through an intuitive user interaction, Ul-Cluster solves the problem mentioned above, and yields better clustering results.

Regiospecific orientation of single-chain antibody and atomic force microscope (AFM) images

An antibody containing a genetically engineered lipid group at the N-terminus and a hexahistidinyl tag at the C-terminus (Lpp-scFv-His6) was immobilized in an oriented manner on the surface of liposomes. Liposomes, consisting of antibody and phosphatidylcholine, have been prepared and imaged by AFM. For AFM visualization, the resulting liposomes were bound on the surface of mica by two different mechanisms. The histidine tags present in the antibody molecules of the immunoliposome were anchored to the NiCl2 treated mica surface. Alternatively, the immunoliposomes were immunochemically bound on antigen-coated mica surface. Both approaches yielded liposomes which were clearly imaged without damage by AFM in ambient condition.