Translate this page to:
In JoVE (3)
- DNA Extraction from 0.22 μM Sterivex Filters and Cesium Chloride Density Gradient Centrifugation
- Large Insert Environmental Genomic Library Production
- Extraction of High Molecular Weight Genomic DNA from Soils and Sediments
Other Publications (21)
- Biochemical and Biophysical Research Communications
- FEMS Microbiology Letters
- Journal of Biomolecular NMR
- Journal of the American Chemical Society
- Journal of the American Chemical Society
- FEMS Microbiology Letters
- Mycological Research
- Journal of Structural and Functional Genomics
- Biochimica Et Biophysica Acta
- Science (New York, N.Y.)
- Protein Science : a Publication of the Protein Society
- Environmental Microbiology
- IEEE Transactions on Neural Systems and Rehabilitation Engineering : a Publication of the IEEE Engineering in Medicine and Biology Society
- ACS Nano
- The Journal of Biological Chemistry
- IEEE Transactions on Signal Processing : a Publication of the IEEE Signal Processing Society
- Proceedings of the National Academy of Sciences of the United States of America
- Journal of Physical Activity & Health
- Conference Proceedings : ... Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Conference
Articles by Sangwon Lee in JoVE
DNA Extraction from 0.22 μM Sterivex Filters and Cesium Chloride Density Gradient Centrifugation
Jody J. Wright, Sangwon Lee, Elena Zaikova, David A. Walsh, Steven J. Hallam
Department of Microbiology and Immunology, University of British Columbia - UBC
We describe a method for extraction of high molecular weight genomic DNA from planktonic biomass concentrated on 0.22 μm Sterivex filters, followed by cesium chloride density gradient centrifugation for purification.
Large Insert Environmental Genomic Library Production
Marcus Taupp, Sangwon Lee, Alyse Hawley, Jinshu Yang, Steven J. Hallam
Department of Microbiology and Immunology, University of British Columbia - UBC
Construction of a fosmid library with environmental genomic DNA isolated from the vertical depth continuum of a seasonally hypoxic fjord is described. The resulting clone library is picked into 384-well plates and archived for downstream sequencing and functional screening by the application of an automated colony picking system.
Extraction of High Molecular Weight Genomic DNA from Soils and Sediments
Sangwon Lee, Steven J. Hallam
Department of Microbiology and Immunology, University of British Columbia - UBC
A methodology to isolate high molecular weight and high quality genomic DNA from soil microbial community is described.
Other articles by Sangwon Lee on PubMed
Biochemical and Biophysical Research Communications. Feb, 2002 | Pubmed ID: 11829477
Aminoacyl-tRNA synthetases (ARSs) ligate amino acids to their cognate tRNAs. It has been suggested that mammalian ARSs are linked to the EF-1 complex for efficient channeling of aminoacyl tRNAs to ribosome. Here we systemically investigated possible interactions between human ARSs and the subunits of EF-1 (alpha, beta, gamma, and delta) using a yeast two-hybrid assay. Among the 80 tested pairs, leucyl- and histidyl-tRNA synthetases were found to make strong and specific interaction with the EF-1gamma and beta while glu-proly-, glutaminyl-, alanyl-, aspartyl-, lysyl-, phenylalanyl-, glycyl-, and tryptophanyl-tRNA synthetases showed moderate interactions with the different EF-1 subunits. The interactions of leucyl- and histidyl-tRNA synthetase with the EF-1 complex were confirmed by immunoprecipitation and in vitro pull-down experiments. Interestingly, the aminoacylation activities of these two enzymes, but not other ARSs, were stimulated by the cofactor of EF-1, GTP. These data suggest that a systematic interaction network may exist between mammalian ARSs and EF-1 subunits probably to enhance the efficiency of in vivo protein synthesis.
FEMS Microbiology Letters. May, 2003 | Pubmed ID: 12770706
Two synonymous sapstain species, Ophiostoma montium and Ophiostoma ips, which are vectored by Dendroctonus ponderosae and various bark beetles, respectively, were differentiated into separate species using growth and molecular characteristics. Analysis of 32 isolates of the two species from different countries showed that O. ips was able to grow at 35 degrees C while O. montium was not. This growth-based differentiation was strongly supported by sequence data for the internal transcribed spacer (ITS), 5.8S and partial 28S rDNA, and the beta-tubulin genes. The beta-tubulin gene sequence data indicate that the two species can easily be differentiated with a single polymerase chain reaction (PCR) assay.
Journal of Biomolecular NMR. Aug, 2003 | Pubmed ID: 12815259
Three solution NMR experiments on a uniformly (15)N labeled membrane protein in micelles provide sufficient information to describe the structure, topology, and dynamics of its helices, as well as additional information that characterizes the principal features of residues in terminal and inter-helical loop regions. The backbone amide resonances are assigned with an HMQC-NOESY experiment and the backbone dynamics are characterized by a (1)H-(15)N heteronuclear NOE experiment, which clearly distinguishes between the structured helical residues and the more mobile residues in the terminal and interhelical loop regions of the protein. The structure and topology of the helices are described by Dipolar waves and PISA wheels derived from experimental measurements of residual dipolar couplings (RDCs) and residual chemical shift anisotropies (RCSAs). The results show that the membrane-bound form of Pf1 coat protein has a 20-residue trans-membrane hydrophobic helix with an orientation that differs by about 90 degrees from that of an 8-residue amphipathic helix. This combination of three-experiments that yields Dipolar waves and PISA wheels has the potential to contribute to high-throughput structural characterizations of membrane proteins.
Journal of the American Chemical Society. Jul, 2003 | Pubmed ID: 12862490
Dipolar waves describe the structure and topology of helices in membrane proteins. The fit of sinusoids with the 3.6 residues per turn period of ideal alpha-helices to experimental measurements of dipolar couplings as a function of residue number makes it possible to simultaneously identify the residues in the helices, detect kinks or curvature in the helices, and determine the absolute rotations and orientations of helices in completely aligned bilayer samples and relative rotations and orientations of helices in a common molecular frame in weakly aligned micelle samples. Since as much as 80% of the structured residues in a membrane protein are in helices, the analysis of dipolar waves provides a significant step toward structure determination of helical membrane proteins by NMR spectroscopy.
Functional Characterization and NMR Spectroscopy on Full-length Vpu from HIV-1 Prepared by Total Chemical Synthesis
Journal of the American Chemical Society. Mar, 2004 | Pubmed ID: 14982452
Vpu is an 81-residue integral membrane protein encoded in the HIV-1 genome that is of considerable interest because it plays important roles in the release of virus particles from infected cells and in the degradation of the cellular receptor. We report here the total chemical synthesis of full-length Vpu(1-81) as well as a site-specifically (15)N-labeled analogue, Vpu(2-81), using native chemical ligation methodologies and also report a structural and functional comparison of these constructs with recombinant protein obtained via bacterial expression. The structures of the synthetic and expressed polypeptides were similar in lipid micelles using solution NMR spectroscopy. Solid-state NMR spectra of the polypeptides in aligned hydrated lipid bilayers indicated that their overall topologies were also very comparable. Further, the channel activity of the synthetic protein was found to be analogous to that previously characterized for the recombinant protein. We have thus demonstrated that using solid phase peptide synthesis and chemical ligation it is feasible to obtain large quantities of a purified and homogeneous membrane protein in a structurally and functionally relevant form for future structural and characterization studies.
Multigene Phylogenies of Ophiostoma Clavigerum and Closely Related Species from Bark Beetle-attacked Pinus in North America
FEMS Microbiology Letters. Aug, 2004 | Pubmed ID: 15268942
Leptographium pyrinum, Leptographium terebrantis, Ophiostoma aureum, Ophiostoma clavigerum, and Ophiostoma robustum are very similar in morphology, host trees choice, and the way they are disseminated by bark beetles. Their phylogenetic relationships were clarified using rDNA and protein coding genes including actin, beta-tubulin, and translation elongation factor-1alpha. Protein coding gene trees showed better resolution than the rDNA tree, which generated three clades: O. clavigerum, L. terebrantis/L. pyrinum, and O. robustum/O. aureum. A combined gene phylogenetic tree, which was supported by high bootstrap values, showed that O. aureum, L. pyrinum, O. robustum, and O. clavigerum each formed distinct clades while L. terebrantis was paraphyletic to O. clavigerum. The higher variability of the protein coding genes and the congruity in their phylogenetic results suggested that these genes may be better markers for identifying closely related species. These gene trees have also facilitated the description of the evolutionary relationships among these species.
Leptographium Longiclavatum Sp. Nov., a New Species Associated with the Mountain Pine Beetle, Dendroctonus Ponderosae
Mycological Research. Oct, 2005 | Pubmed ID: 16279410
The mountain pine beetle, Dendroctonus ponderosae, and its fungal associates are devastating the lodgepole pine forests in British Columbia, Canada. During our fungal survey, an unknown Leptographium species has been consistently isolated from both D. ponderosae and infested lodgepole pine (Pinus contorta var. latifolia). This Leptographium species has similar morphology with the Leptographium anamorph of Ophiostoma clavigerum whose association with the D. ponderosae is well known. However, thorough morphological comparisons showed that this fungus is distinct from all the other Leptographium species described in the literature, and especially from O. clavigerum. Comparison of DNA sequences of multiple loci and the profiles by the PCR-RFLP marker also confirmed that this Leptographium species represents an undescribed taxon. Based on its distinct morphological, physiological characteristics and phylogenetic position, we describe it as L. longiclavatum sp. nov.
Comprehensive Evaluation of Solution Nuclear Magnetic Resonance Spectroscopy Sample Preparation for Helical Integral Membrane Proteins
Journal of Structural and Functional Genomics. Mar, 2006 | Pubmed ID: 16850177
The preparation of high quality samples is a critical challenge for the structural characterization of helical integral membrane proteins. Solving the structures of this diverse class of proteins by solution nuclear magnetic resonance spectroscopy (NMR) requires that well-resolved 2D 1H/15N chemical shift correlation spectra be obtained. Acquiring these spectra demands the production of samples with high levels of purity and excellent homogeneity throughout the sample. In addition, high yields of isotopically enriched protein and efficient purification protocols are required. We describe two robust sample preparation methods for preparing high quality, homogeneous samples of helical integral membrane proteins. These sample preparation protocols have been combined with screens for detergents and sample conditions leading to the efficient production of samples suitable for solution NMR spectroscopy. We have examined 18 helical integral membrane proteins, ranging in size from approximately 9 kDa to 29 kDa with 1-4 transmembrane helices, originating from a number of bacterial and viral genomes. 2D 1H/15N chemical shift correlation spectra acquired for each protein demonstrate well-resolved resonances, and >90% detection of the predicted resonances. These results indicate that with proper sample preparation, high quality solution NMR spectra of helical integral membrane proteins can be obtained greatly enhancing the probability for structural characterization of these important proteins.
NMR and Mutagenesis of Human Copper Transporter 1 (hCtr1) Show That Cys-189 is Required for Correct Folding and Dimerization
Biochimica Et Biophysica Acta. Dec, 2007 | Pubmed ID: 17959139
The human high-affinity copper transporter (hCtr1) is a membrane protein that is predicted to have three transmembrane helices and two methionine-rich metal binding motifs. As an oligomeric polytopic membrane protein, hCtr1 is a challenging system for experimental structure determination. The results of an initial application of solution-state NMR methods to a truncated construct containing residues 45-190 in micelles and site-directed mutagenesis of the two cysteine residues demonstrate that Cys-189 but not Cys-161 is essential for both dimer formation and proper folding of the protein.
Science (New York, N.Y.). Feb, 2008 | Pubmed ID: 18218865
We developed a chemical route to produce graphene nanoribbons (GNR) with width below 10 nanometers, as well as single ribbons with varying widths along their lengths or containing lattice-defined graphene junctions for potential molecular electronics. The GNRs were solution-phase-derived, stably suspended in solvents with noncovalent polymer functionalization, and exhibited ultrasmooth edges with possibly well-defined zigzag or armchair-edge structures. Electrical transport experiments showed that, unlike single-walled carbon nanotubes, all of the sub-10-nanometer GNRs produced were semiconductors and afforded graphene field effect transistors with on-off ratios of about 10(7) at room temperature.
Backbone Structure of a Small Helical Integral Membrane Protein: A Unique Structural Characterization
Protein Science : a Publication of the Protein Society. Jan, 2009 | Pubmed ID: 19177358
The structural characterization of small integral membrane proteins pose a significant challenge for structural biology because of the multitude of molecular interactions between the protein and its heterogeneous environment. Here, the three-dimensional backbone structure of Rv1761c from Mycobacterium tuberculosis has been characterized using solution NMR spectroscopy and dodecylphosphocholine (DPC) micelles as a membrane mimetic environment. This 127 residue single transmembrane helix protein has a significant (10 kDa) C-terminal extramembranous domain. Five hundred and ninety distance, backbone dihedral, and orientational restraints were employed resulting in a 1.16 A rmsd backbone structure with a transmembrane domain defined at 0.40 A. The structure determination approach utilized residual dipolar coupling orientation data from partially aligned samples, long-range paramagnetic relaxation enhancement derived distances, and dihedral restraints from chemical shift indices to determine the global fold. This structural model of Rv1761c displays some influences by the membrane mimetic illustrating that the structure of these membrane proteins is dictated by a combination of the amino acid sequence and the protein's environment. These results demonstrate both the efficacy of the structural approach and the necessity to consider the biophysical properties of membrane mimetics when interpreting structural data of integral membrane proteins and, in particular, small integral membrane proteins.
Bacterial, Archaeal and Eukaryal Community Structures Throughout Soil Horizons of Harvested and Naturally Disturbed Forest Stands
Environmental Microbiology. Dec, 2009 | Pubmed ID: 19659501
Disturbances caused by timber harvesting have critical long-term effects on the forest soil microbiota and alter fundamental ecosystem services provided by these communities. This study assessed the effects of organic matter removal and soil compaction on microbial community structures in different soil horizons 13 years after timber harvesting at the long-term soil productivity site at Skulow Lake, British Columbia. A harvested stand was compared with an unmanaged forest stand. Ribosomal intergenic spacer profiles of bacteria, archaea and eukarya indicated significantly different community structures in the upper three soil horizons of the two stands, with differences decreasing with depth. Large-scale sequencing of the ribosomal intergenic spacers coupled to small-subunit ribosomal RNA genes allowed taxonomic identification of major microbial phylotypes affected by harvesting or varying among soil horizons. Actinobacteria and Gemmatimonadetes were the predominant phylotypes in the bacterial profiles, with the relative abundance of these groups highest in the unmanaged stand, particularly in the deeper soil horizons. Predominant eukaryal phylotypes were mainly assigned to known mycorrhizal and saprotrophic species of Basidiomycetes and Ascomycetes. Harvesting affected Basidiomycetes to a minor degree but had stronger effects on some Ascomycetes. Archaeal profiles had low diversity with only a few predominant crenarchaeal phylotypes whose abundance appeared to increase with depth. Detection of these effects 13 years after harvesting may indicate a long-term change in processes mediated by the microbial community with important consequences for forest productivity. These effects warrant more comprehensive investigation of the effects of harvesting on the structure of forest soil microbial communities and the functional consequences.
IEEE Transactions on Neural Systems and Rehabilitation Engineering : a Publication of the IEEE Engineering in Medicine and Biology Society. Aug, 2010 | Pubmed ID: 20699202
A physical activity (PA) recognition algorithm for a wearable wireless sensor network using both ambulatory electrocardiogram (ECG) and accelerometer signals is proposed. First, in the time domain, the cardiac activity mean and the motion artifact noise of the ECG signal are modeled by a Hermite polynomial expansion and principal component analysis, respectively. A set of time domain accelerometer features is also extracted. A support vector machine (SVM) is employed for supervised classification using these time domain features. Second, motivated by their potential for handling convolutional noise, cepstral features extracted from ECG and accelerometer signals based on a frame level analysis are modeled using Gaussian mixture models (GMMs). Third, to reduce the dimension of the tri-axial accelerometer cepstral features which are concatenated and fused at the feature level, heteroscedastic linear discriminant analysis is performed. Finally, to improve the overall recognition performance, fusion of the multimodal (ECG and accelerometer) and multidomain (time domain SVM and cepstral domain GMM) subsystems at the score level is performed. The classification accuracy ranges from 79.3% to 97.3% for various testing scenarios and outperforms the state-of-the-art single accelerometer based PA recognition system by over 24% relative error reduction on our nine-category PA database.
ACS Nano. Dec, 2010 | Pubmed ID: 21121692
Owing to its unique electronic properties, graphene has recently attracted wide attention in both the condensed matter physics and microelectronic device communities. Despite intense interest in this material, an industrially scalable graphene synthesis process remains elusive. Here, we demonstrate a high-throughput, low-temperature, spatially controlled and scalable epitaxial graphene (EG) synthesis technique based on laser-induced surface decomposition of the Si-rich face of a SiC single-crystal. We confirm the formation of EG on SiC as a result of excimer laser irradiation by using reflection high-energy electron diffraction (RHEED), Raman spectroscopy, synchrotron-based X-ray diffraction, transmission electron microscopy (TEM), and scanning tunneling microscopy (STM). Laser fluence controls the thickness of the graphene film down to a single monolayer. Laser-synthesized graphene does not display some of the structural characteristics observed in EG grown by conventional thermal decomposition on SiC (0001), such as Bernal stacking and surface reconstruction of the underlying SiC surface.
Biochemistry. Jun, 2011 | Pubmed ID: 21574595
Amyloid precursor protein (APP) is genetically linked to Alzheimer's disease. APP is a type I membrane protein, and its oligomeric structure is potentially important because this property may play a role in its function or affect the processing of the precursor by the secretases to generate amyloid β-peptide. Several independent studies have shown that APP can form dimers in the cell, but how it dimerizes remains controversial. At least three regions of the precursor, including a centrally located and conserved domain called E2, have been proposed to contribute to dimerization. Here we report two new crystal structures of E2, one from APP and the other from APLP1, a mammalian APP homologue. Comparison with an earlier APP structure, which was determined in a different space group, shows that the E2 domains share a conserved and antiparallel mode of dimerization. Biophysical measurements in solution show that heparin binding induces E2 dimerization. The 2.1 Å resolution electron density map also reveals phosphate ions that are bound to the protein surface. Mutational analysis shows that protein residues interacting with the phosphate ions are also involved in heparin binding. The locations of two of these residues, Arg-369 and His-433, at the dimeric interface suggest a mechanism for heparin-induced protein dimerization.
Crystal Structure of the E2 Domain of Amyloid Precursor Protein-like Protein 1 in Complex with Sucrose Octasulfate
The Journal of Biological Chemistry. Aug, 2011 | Pubmed ID: 21715329
Missense mutations in the amyloid precursor protein (APP) gene can cause familial Alzheimer disease. It is thought that APP and APP-like proteins (APLPs) may play a role in adhesion and signal transduction because their ectodomains interact with components of the extracellular matrix. Heparin binding induces dimerization of APP and APLPs. To help explain how these proteins interact with heparin, we have determined the crystal structure of the E2 domain of APLP1 in complex with sucrose octasulfate (SOS). A total of three SOS molecules are bound to the E2 dimer. Two SOSs are bound inside a narrow intersubdomain groove, and the third SOS is bound near the two-fold axis of the protein. Mutational analyses show that most residues interacting with SOS also contribute to heparin binding, although in varying degrees; a deep pocket, defined by His-376, Lys-422, and Arg-429, and an interfacial site between Lys-314 and its symmetry mate are most important in the binding of the negatively charged polysaccharide. Comparison with a lower resolution APP structure shows that all key heparin binding residues are conserved and identically positioned, suggesting that APLP1 and APP may bind heparin similarly. In transfected HEK-293 cells, mutating residues responsible for heparin binding causes little change in the proteolysis of APP by the secretases. However, mutating a pair of conserved basic residues (equivalent to Arg-414 and Arg-415 of APLP1) immediately adjacent to the heparin binding site affects both the maturation and the processing of APP.
Nature. Jul, 2011 | Pubmed ID: 21765428
The GXGD proteases are polytopic membrane proteins with catalytic activities against membrane-spanning substrates that require a pair of aspartyl residues. Representative members of the family include preflagellin peptidase, type 4 prepilin peptidase, presenilin and signal peptide peptidase. Many GXGD proteases are important in medicine. For example, type 4 prepilin peptidase may contribute to bacterial pathogenesis, and mutations in presenilin are associated with Alzheimer's disease. As yet, there is no atomic-resolution structure in this protease family. Here we report the crystal structure of FlaK, a preflagellin peptidase from Methanococcus maripaludis, solved at 3.6 Å resolution. The structure contains six transmembrane helices. The GXGD motif and a short transmembrane helix, helix 4, are positioned at the centre, surrounded by other transmembrane helices. The crystal structure indicates that the protease must undergo conformational changes to bring the GXGD motif and a second essential aspartyl residue from transmembrane helix 1 into close proximity for catalysis. A comparison of the crystal structure with models of presenilin derived from biochemical analysis reveals three common transmembrane segments that are similarly arranged around the active site. This observation reinforces the idea that the prokaryotic and human proteases are evolutionarily related. The crystal structure presented here provides a framework for understanding the mechanism of the GXGD proteases, and may facilitate the rational design of inhibitors that target specific members of the family.
IEEE Transactions on Signal Processing : a Publication of the IEEE Signal Processing Society. 2011 | Pubmed ID: 21796237
The optimal allocation of samples for physical activity detection in a wireless body area network for health-monitoring is considered. The number of biometric samples collected at the mobile device fusion center, from both device-internal and external Bluetooth heterogeneous sensors, is optimized to minimize the transmission power for a fixed number of samples, and to meet a performance requirement defined using the probability of misclassification between multiple hypotheses. A filter-based feature selection method determines an optimal feature set for classification, and a correlated Gaussian model is considered. Using experimental data from overweight adolescent subjects, it is found that allocating a greater proportion of samples to sensors which better discriminate between certain activity levels can result in either a lower probability of error or energy-savings ranging from 18% to 22%, in comparison to equal allocation of samples. The current activity of the subjects and the performance requirements do not significantly affect the optimal allocation, but employing personalized models results in improved energy-efficiency. As the number of samples is an integer, an exhaustive search to determine the optimal allocation is typical, but computationally expensive. To this end, an alternate, continuous-valued vector optimization is derived which yields approximately optimal allocations and can be implemented on the mobile fusion center due to its significantly lower complexity.
Crystal Structure of Amyloid Precursor-like Protein 1 and Heparin Complex Suggests a Dual Role of Heparin in E2 Dimerization
Proceedings of the National Academy of Sciences of the United States of America. Sep, 2011 | Pubmed ID: 21930949
Mutations in amyloid precursor protein (APP) are associated with familial Alzheimer's disease. Recent development suggests that homo- and heterodimerization of APP and APP-like proteins (APLPs), which are enhanced by heparan sulfate binding, may play a role in signal transduction and cell adhesion. Despite efforts to model heparin binding based on known apo crystal structures, the mechanism of heparin-induced APP/APLP dimerization has not been established experimentally. Here we report the crystal structure of a complex between heparin and the E2 domain of APLP1, which harbors the conserved high affinity heparin binding site of the full-length molecule. Within the asymmetric E2:heparin complex, the polysaccharide is snugly bound inside a narrow groove between the two helical subdomains of one protein protomer. The nonreducing end of the sugar is positioned near the protein's 2-fold axis, making contacts with basic residues from the second protomer. The inability of the E2 dimer to accommodate two heparin molecules near its symmetry axis explains the observed 21 binding stoichiometry, which is confirmed by isothermal titration calorimetric experiment carried out in solution. We also show that, at high concentrations, heparin can destabilize E2 dimer, probably by forcing into the unoccupied binding site observed in the 21 complex. The binding model suggested by the crystal structure may facilitate the design of heparin mimetics that are capable of modulating APP dimerization in cells.
Journal of Physical Activity & Health. May, 2011 | Pubmed ID: 21934162
BACKGROUND: KNOWME Networks is a wireless body area network with two tri-axial accelerometers, a heart rate monitor, and mobile phone that acts as the data collection hub. One function of KNOWME Networks is to detect physical activity (PA) in overweight Hispanic youth. The purpose of this study was to evaluate the in-lab recognition accuracy of KNOWME. METHODS: Twenty overweight Hispanic participants (10 males; age 14.6±1.8 years), underwent four data collection sessions consisting of nine activities/session: lying down, sitting, sitting fidgeting, standing, standing fidgeting, standing playing an active video game, slow walking, brisk walking, and running. Data was used to train activity recognition models. The accuracy of personalized and generalized models is reported. RESULTS: Overall accuracy for personalized models was 84%. The most accurately detected activity was running (96%). The models had difficulty distinguishing between the static and fidgeting categories of sitting and standing. When static and fidgeting activity categories were collapsed, the overall accuracy improved to 94%. Personalized models demonstrated higher accuracy than generalized models. CONCLUSIONS: KNOWME Networks can accurately detect a range of activities. KNOWME has the ability to collect and process data in real-time, building the foundation for tailored, real-time interventions to increase PA or decrease sedentary time.
Modeling High-level Descriptions of Real-life Physical Activities Using Latent Topic Modeling of Multimodal Sensor Signals
Conference Proceedings : ... Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Conference. Aug, 2011 | Pubmed ID: 22255715
We propose a new methodology to model high-level descriptions of physical activities using multimodal sensor signals (ambulatory electrocardiogram (ECG) and accelerometer signals) obtained by a wearable wireless sensor network. We introduce a two-step strategy where the first step estimates likelihood scores over the low-level descriptions of physical activities such as walking or sitting directly from sensor signals and the second step infers the high-level description based on the estimated low-level description scores. Assuming that a high-level description of a certain physical activity may consist of multiple low-level physical activities and a low-level physical activity can be observed in multiple high-level descriptions of physical activities, we introduce the statistical concept of latent topics in physical activities to model the high-level status with low-level descriptions. With an unsupervised approach using a database from unconstrained free-living settings, we show promising results in modeling high-level descriptions of physical activities.